主讲嘉宾：ROGER B. DANNENBERG教授
主持人：李小兵教授
时间：2022年4月25日 20:00-21:30
题目：算法作曲：从流行歌曲到创造性的信号处理
主办：音乐人工智能与音乐信息科技系
入场方式
ZOOM会议号：978 438 6825
入会密码：8888
Title:
Algorithmic Composition: From Pop Songs to Creative Signal Processing
算法作曲：从流行歌曲到创造性的信号处理
Topic:
Procedural or algorithmic methods, AI, and Machine Learning are all interesting ways to produce music. Algorithmic composition began in the 1950's and continues to offer opportunities for composers to explore new musical ideas. More recently, AI and Machine Learning has received much attention and shown great promise, but results so far have not delivered the controllability or even quality of many previous works. I will describe recent work on automatic generation of Pop Songs using statistical models and explain various ways that hierarchical structure can be used to guide the creation of songs. Our most recent work attempts to replace some of the algorithmic or rule-based methods with machine learning, although results are not necessarily better. Rather than generate traditional notes and rhythms, we can turn to computation to control real-time musical signal processing. Simple effects such as delays and pitch shifters can create novel sounds and play an important role in composition when they are continuously controlled by computer. I will present some recent explorations of "algorithmic signal processing" for live performance.
程式化作曲、算法作曲、人工智能和机器学习作曲都是非常有趣的音乐创作方式。算法作曲诞生于1950年代，至今仍在为作曲家们创造探索新乐思的机会。近年来，人工智能和机器学习备受关注，并且有着很好的前景，但目前为止，生成结果尚不具有可控制性，甚至其质量也无法与先前的作品相提并论。本次报告将首先介绍近来利用统计模型生成流行歌曲的工作，并阐述多种利用分层结构指导歌曲生成的方法。近期音乐生成方面的工作，都在尝试用机器学习方法替代一部分基于算法或规则的生成方法，尽管其结果并不一定更好。除了按照传统方法生成音符和节奏，我们还能够利用计算方法来控制实时的音乐信号处理，产生一些比较简单的效果，例如延迟和变调，这样可以创造出新颖的声音，并且当我们持续地用计算机控制它们时，这些声音将在音乐创作中起到重要作用，在本次报告中也将展示近来一些针对现场表演中算法信号处理的探索。
Biography:
Roger B. Dannenberg is a Professor Emeritus of Computer Science, Art & Music. He received a Ph.D. in Computer Science from Carnegie Mellon University in 1982. He is internationally known for his research in the field of computer music. His current work includes live music performance with artificial computer musicians, interactive media, high-level languages for sound synthesis, and computer support of curriculum design. Prof. Dannenberg is also a trumpet player and composer. He has performed in concert halls ranging from the historic Apollo Theater in Harlem to the modern Espace de Projection at IRCAM in Paris. Besides numerous compositions for musicians and interactive electronics, Dannenberg co-composed the opera La Mare dels Peixos with Jorge Sastre.
Roger B. Dannenberg 现任卡内基梅隆大学计算机科学、艺术和音乐的名誉教授。他于1982 年获得卡内基梅隆大学计算机科学博士学位，在计算机音乐领域的研究享誉国际。他目前的研究工作包括“计算机音乐家”的现场表演、交互式媒体、声音合成的高级语言以及课程设计的计算机支持。Dannenberg 教授也是一位小号手和作曲家。他曾在许多音乐厅演出，从哈莱姆历史悠久的阿波罗剧院到巴黎 IRCAM 的现代 Espace de Projection。除了为音乐家与互动电子设备创作众多作品外，Dannenberg 还与 Jorge Sastre 共同创作了歌剧 La Mare dels Peixos。
供稿：音乐人工智能与音乐信息科技系
文：张渊、周子雅
