Audio signal processing and refinement is an important part of a content-based music information retrieval system. As the our repertoire of techniques becomes more varied, there are greater requirements of computation power. Distributed storage techniques have become widespread and almost invisible with the advent of file-sharing systems, on-line digital music stores and on line storage services. Even discounting data with potential copyright entanglements, there is a vast amount that is ripe for analysis, and thus parallelised and distributed processing techniques seem increasingly appropriate. Existing frameworks are already capable of a significant amount of audio analysis for music information retrieval. However they are by and large ignorant of distribution and parallelisation. There are middleware libraries to help with aspects of distributed computing, but combining the two can be cumbersome and inefficient. This paper provides a brief description of a software framework that can process audio in a scalable and distributed manner: Geddei. The paper then takes an interesting and relevant signal analysis task often used for music information retrieval and implements it under the Geddei framework. The ease of use is discussed and various measurements taken of Geddei, both in comparison to itself under different circumstances, and ‘reference code’ that was used in a previous study. We discuss the problems with the distribution of the task with Geddei and offer some possible solutions.
[1]
Perry R. Cook,et al.
Music, cognition, and computerized sound: an introduction to psychoacoustics
,
1999
.
[2]
Andreas Rauber,et al.
Automatically Analyzing and Organizing Music Archives
,
2001,
ECDL.
[3]
J. Stephen Downie,et al.
Report on the panels and workshops of the music information retrieval (MIR) and music digital library (MDL) evaluation frameworks project
,
2003,
SIGF.
[4]
Simon O'Keefe,et al.
Quantitative comparisons into content-based music recognition with the self organising map
,
2003,
ISMIR.
[5]
George Tzanetakis,et al.
MARSYAS: a framework for audio analysis
,
1999,
Organised Sound.
[6]
Jonathan Foote,et al.
Audio Retrieval by Rhythmic Similarity
,
2002,
ISMIR.
[7]
George Tzanetakis,et al.
Automatic Musical Genre Classification of Audio Signals
,
2001,
ISMIR.
[8]
George Tzanetakis,et al.
Audio Information Retrieval (AIR) Tools
,
2000,
ISMIR.
[9]
George Tzanetakis,et al.
Musical genre classification of audio signals
,
2002,
IEEE Trans. Speech Audio Process..