The Music Features Ontology Revision: 1.00 This ontology express the most common things you may extract automatically from a musical signal, by sub-classing some of the terms defined in the Music Ontology. $Date: 2007/02/28 11:26:48 $ Output of a segmentation process (ie. an algorithm taking as an input the musical signal, and outputting a list of segments). More specific type of segment. Here our segmentation algorithm is trying to get to more specific types of segments, related to the structure of the underlying musical piece (intro/verse/chorus/bridge/outro/...). Subclasses of this class should represent particular segment types (maybe labelled) outputted by our process. Instances should represent occurences of these segment types. Output a key detection process. In case of a "clean cut", instances of such events should have one factor: the detected Key (http://purl.org/NET/c4dm/keys.owl#Key). In other cases, we can put several keys as factors of such events, and attach to them a confidence value (using event decomposition). Output of an onset detection process. The associated time object should be an Instant. A tonal onset - we can attach a Note as a factor of such events. A non-tonal onset. Use this property to attach a confidence value to an event. This property can be used to express things such as "the probability of having this key at this time is ... ".