縮編式自動編曲之研究

Automatic Music Arrangement by Score Reduction

邱士銓

指導教授 ： 黃俊龍 ； 沈錳坤

國立交通大學/資訊學院/資訊科學與工程研究所/博士(2011年)

https://doi.org/10.6842/NCTU.2011.00096

摘要

樂譜縮編(Score Reduction)是一個透過縮編樂譜來達成為單一樂器編曲的過程。在本篇論文中，我們提出了一個使用樂譜縮編方法的編曲架構，此架構可以自動化為一樣樂器進行編曲。根據樂譜縮編的方法，我們要盡可能地包含原曲的每一個部份，並同時滿足目標樂器可彈奏性的限制，使得編曲出來的音樂聽起來和原曲相同。在本架構的第一個步驟中，我們分析原始曲目的音樂編曲元素(Arrangement Element)。接著，將音樂中的每個樂句辨識出來，並且根據音樂編曲元素分析的結果與樂句的特性，來分配每個樂句的重要程度。最後，我們將音樂編曲轉換成一個最佳化的問題並設計一個演算法來解決這個問題。藉由挑選適當的樂句並且同時考量目標樂器的可彈奏性，來完成編曲。在實驗中，我們使用這個編曲架構來實作一個鋼琴編曲的系統。許多的實驗被設計來評估我們系統所產生出來的音樂。為了避免主觀的影響，我們採用了一個類似圖林測試(Turing Test)的方法來評估這個系統的好壞。實驗的結果證明我們的系統有能力編寫出具品質且可彈奏的曲子。
此外，為了捕捉到原始音樂的特色，我們介紹了一種新的樣式－多音重覆樣式，並提出兩個演算法－A-PRPD (Apriori-based Polyphonic Repeating Pattern Discovery) and D-PRPD (Depth-first-search based Polyphonic Repeating Pattern Discovery)，從原始音樂中探勘多音重覆樣式。再者，我們設計了位元方法(bit approach)用於我們所提出的兩個演算法上加速運算。實驗結果顯示，我們提出的演算法是有效率與效果的，並且D-PRPD演算法加上位元方法在大多數的情況下是最有效率的演算法。此探勘出的重覆樣式可以被使用在此音樂編曲架構中的功能性分配的步驟中，使得具有原始音樂特色的樂句較容易被挑選到。

關鍵字

縮編式編曲 ； 自動音樂編曲 ； 鋼琴縮編 ； 電腦音樂

並列摘要

Score reduction is a process that arranges music for a target instrument by reducing original music. In this dissertation we present a music arrangement framework that uses score reduction to automatically arrange music for a target instrument. According to the approach of score reduction, the goal is to include as many important parts of the original music as possible within the constraint of the target instrument so that the arranged version is similar to the original. In our proposed framework, the original music is first analyzed to determine the type of arrangement element of each section. Then, the phrases are identified and each is assigned a utility according to its type of arrangement element. For a set of utility-assigned phrases, we finally transform the music arrangement into an optimization problem and propose a phrase selection algorithm to solve it. The music is arranged by selecting appropriate phrases satisfying the playability constraints of a target instrument.　Using the proposed framework, we implement a music arrangement system for the piano in our experiments. Several experiments were conducted to evaluate our system. To avoid subjective opinions, one approach of the experiments similar to Turing-test is used to evaluate the quality of the music arranged by our system. The experimental results show that our system is able to create viable music for the piano.
To capture the characteristics of the music for enhancing the proposed music arrangement framework, we introduce a new type of repeating patterns, polyphonic repeating pattern and propose algorithms, A-PRPD (Apriori-based Polyphonic Repeating Pattern Discovery) and D-PRPD (Depth-first-search based Polyphonic Repeating Pattern Discovery), to discover them from music data. Furthermore, a bit-string approach is developed for improving the efficiency of both proposed algorithms. Experimental results show that the proposed algorithms are both effective and efficient for mining polyphonic repeating patterns from synthetic music data and real data, and D-PRPD with bit-string approach is the most efficient approach in most cases. The discovered polyphonic repeating patterns can be used to enhance in the phrase identification and utility assignment phase of our proposed framework such that the phrases with music characteristics will be easy to be selected.

並列關鍵字

score reduction ； automatic music arrangement ； piano reduction ； computer music

參考文獻

[13] D. Conklin and C. Anagnostopoulou, “Representation and Discovery of Multiple Viewpoint Patterns,” In Proc. of International Conference on Computer Music, (ICMC'01), 2001.

[1] E. F. Adebiyi, T. Jiang and M. Kaufmann, “An Efficient Algorithm for Finding Short Approximate Non-tandem Repeats,” Bioinformatics, Vol. 17, No. 1, pp. S5-S12, 2001.

[2] R. Agrawal and R. Srikant, “Fast Algorithms for Mining Association Rules,” In Proc. of International Conference on Very Large Data Bases, (VLDB'94), 1994.

[4] G. Benson, “A Space Efficient Algorithm for Finding the Best Non-overlapping Alignment Score,” Theoretical Computer Science, Vol. 145, No. 1&2, pp. 357-369, 1995.

[6] B. Boser, I. Guyon and V. Vapnik, “A Training Algorithm for Optimal Margin Classifiers,” In Proc. of the Fifth Annual Computational Learning Theory, 1992.