A music information retrieval system can extract information that arises from how various sound sources are panned between channels during the mixing and recording process. The authors propose augmenting standard audio features, which are based on the source music, with one of two methods for extracting panning and contrast features. These additional features provide statistically important information for nontrivial audio classifications tasks. Traditional classifications focus on information about pitch, rhythm, and timbre. Other types of mixing parameters are proposed for future work.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.