General characteristics analysis of Chinese folk songs based on layered stabilities detection (LSD) audio segmentation algorithm

This paper discusses a method to automatically analyze the general characteristics of Chinese folk songs. This not only makes it possible to find general characteristics from a large number of folk songs, but it also provides people with a more profound understanding of the creation of them. We use the styles of folk songs' music structure types we proposed in each region to study the general characteristics. The process consists of three steps: first, segment each folk song into clips based on LSD audio segmentation algorithm we first proposed. Then, music structure annotation to these clips. Finally, make statistics on the styles of each folk song's music structure types and analyze their general characteristics. Experiments show that it is feasible to automatically analyze the general characteristics of folk songs based on the styles of music structure types we proposed. The general characteristics of the folk songs in three regions are based on the reality that all music structure types and styles have similar ratios, the Coordinate Structure has the most, and the Cyclotron Structure has the least.