複数旋律音楽に対する演奏表情付けモデルの構築 A Performance Rendering Model for Polyphrase Ensemble

この論文にアクセスする

この論文をさがす

著者

抄録

演奏表情付けシステムの音楽表現能力は近年おおいに向上してきたが,複数の旋律からなる楽曲の演奏表現についてはあまり考慮されていなかった.本論文では,ユーザの介在を前提とした演奏デザイン支援を行うための複数旋律の楽曲に対する基礎的な演奏表現モデルPop-E(PolyphraseEnsemble)を提案し,その実装例について述べる.Pop-E では,各旋律の演奏表現を実施するにあたり,まず,声部別に演奏ルールを適用する.この結果,複数の声部間で演奏に必要な占有時間が異なり,声部間で発音のタイミングを揃えたい時刻にもずれが生じてしまう.この対策として,グループ構造と演奏上のアテンションパートを手がかりにして必要に応じて同期させる処理手法を導入している.本モデルに基づいて生成した「幻想即興曲」は,NIME-Rencon における聴き比べコンテストにおいて第1 位を受賞した.3 人のピアニストの演奏を対象としたモデルの再構築実験では,音量よりも音長(テンポ)に対する高い再現性が確認され,Pop-E によって,少数のルール群で人間に近い演奏を生成しうる可能性が示唆された.  付録:<a href="http://www.ipsj.or.jp/08editt/contents/JNL4801/index.html#24"target="_brank">http://www.ipsj.or.jp/08editt/contents/JNL4801/index.html#24</a>While the expressive capability of music performance rendering systems greatly improved in the last few years, the rendition of multi-part ensembles still requires more detailed consideration. This paper presents the design and implementation of Pop-E (Polyphrase Ensemble), a performance rendering model for multi-part music that supports performance design with intervention from the user. Pop-E .rst generates expressive phrasing for individual parts, assigning independent timings for notes in each part. The notes among different parts are synchronized to generate a coherent ensemble, using a method based on grouping structure and "attentive part" markings. The performance of "Fantaisie-Impromptu" (Chopin, middle part) generated by Pop-E received .rst prize in the hearing contest at NIME-Rencon. The paper also presents evaluation of the descriptive capabilities of Pop-E, based on the resynthesis of performance by three pianists. The results of this experiment, as well as the results from hearing experiments, suggest that Pop-E is capable of generating performer-quality renditions from a small set of performance rules.appendices:<a href="http://www.ipsj.or.jp/08editt/contents/JNL4801/index.html#24"target="_brank">http://www.ipsj.or.jp/08editt/contents/JNL4801/index.html#24</a>

While the expressive capability of music performance rendering systems greatly improved in the last few years, the rendition of multi-part ensembles still requires more detailed consideration. This paper presents the design and implementation of Pop-E (Polyphrase Ensemble), a performance rendering model for multi-part music that supports performance design with intervention from the user. Pop-E first generates expressive phrasing for individual parts, assigning independent timings for notes in each part. The notes among different parts are synchronized to generate a coherent ensemble, using a method based on grouping structure and "attentive part" markings. The performance of "Fantaisie-Impromptu" (Chopin, middle part) generated by Pop-E received first prize in the hearing contest at NIME-Rencon. The paper also presents evaluation of the descriptive capabilities of Pop-E, based on the resynthesis of performance by three pianists. The results of this experiment, as well as the results from hearing experiments, suggest that Pop-E is capable of generating performer-quality renditions from a small set of performance rules.

収録刊行物

  • 情報処理学会論文誌

    情報処理学会論文誌 48(1), 248-257, 2007-01-15

    一般社団法人情報処理学会

参考文献:  22件中 1-22件 を表示

被引用文献:  4件中 1-4件 を表示

各種コード

  • NII論文ID(NAID)
    110006152201
  • NII書誌ID(NCID)
    AN00116647
  • 本文言語コード
    JPN
  • 資料種別
    Journal Article
  • ISSN
    1882-7764
  • NDL 記事登録ID
    8649916
  • NDL 雑誌分類
    ZM13(科学技術--科学技術一般--データ処理・計算機)
  • NDL 請求記号
    Z14-741
  • データ提供元
    CJP書誌  CJP引用  NDL  NII-ELS  IPSJ 
ページトップへ