摘要
主要研究如何将现代乐谱自动翻译成古琴减字谱。古琴减字谱和现代乐谱不同之处在于它是一种记指法的谱子。本文先对曲调进行分析,选择古琴适合的调与之相对应,然后将乐曲分成虚音、主干部分、半虚半实音进行处理。虚音部分采用最大匹配法进行翻译;主干部分采用基于统计的方法、模拟退火算法以及Viterbi算法进行翻译;半虚半实部分采用基于规则的方法进行分析处理。最后得到符合古琴指法要求的古琴减字谱序列。
The paper focus on translation from modern western score to Guqin Jianzi notation. Guqin Jianzi notation is very different from western score which tells fingers to be placed instead of defining an absolute pitch.According to the characteristics of Guqin music, we devide the modern music into three parts: imaginary tone, main part and semi-real part. We apply Maximum matching algorithm and rule-based method to deal with imaginary tone and semi-real part respectively. Statistical method, simulated annealing and viterbi algorithm are combined together to decide the translation of main part. The sequence of Guqin Jianzi characters we obtained is in accordance with Guqin’s fingering rules.
出处
《心智与计算》
2010年第1期45-54,共10页
Mind and Computation