BDB Bertsolaritzaren datu-basea

The steps for melodic description can be summarized as follows. We first compute a set of low- level features from the audio signal, including energy and fundamental frequency. Next, since the tuning reference is unknown, we estimate it with a method based on the equal-tempered chromatic scale. The audio signal is then segmented into notes using a dynamic programming algorithm whose cost functions take into account pitch error, energy variations and note durations. The obtained notes are labeled in terms of pitch and duration. Finally, vibrato depth and rate descriptors are computed for each note. We have worked with a music collection of 135 monophonic pieces corresponding to deblas, martinetes and tonás. These categories of flamenco styles (called palos) are traditionally sung a cappella or in some cases with some percussion. Results. The tool provides a way to visualize different aspects of the analyzed performances, as the instantaneous evolution of energy and pitch, the sequence of notes and some expressive aspects (modulations and transitions). Conclusions. The work presented here provides a set of techniques for the automatic melodic description of flamenco singing. Quantitative evaluation is a difficult task, as it is necessary to have a set of manually annotated performances by expert flamencologists. This is an ongoing work, although some informal tests have proven the usefulness of the approach. Actually, the proposed tool is being used in the context of a research project on comparative analysis of a cappella flamenco styles (Escobar et alt. 2007). Implications. These techniques can support research on comparative analysis of singers and styles, and they provide an automatic and objective way to transcribe flamenco performances.