This text deals with the difficult task of notating timbre by addressing how it can be classified, synthesised, recognised and related to visual correspondences, and then looking at the relevance of these topics for notational purposes. Timbre is understood as dependent on both spectral and time-dependent features that can be notated in ways that make sense in relation to both perception and acoustics. This is achieved by taking the starting point in Lasse Thoresen’s spectromorphological analysis. Symbols originally developed for perception-based analysis are adapted for use over a hybrid spectrum-staff system to indicate the spectral qualities of timbre. To test the system, it was used to transcribe excerpts of three classic electroacoustic music works. In addition to the benefit of being able to compare the three excerpts transcribed with the same system, there is the advantage that the visual representation is based on spectral measurable qualities in the music. The notation system’s intuitiveness was also explored in listening tests, showing that it was possible to understand spectral notation symbols placed over a staff system, particularly for examples with two sound objects instead of one.

本文通过解决如何对音色进行分类、合成、识别以及如何与视觉对应相关联来处理音色标注的艰巨任务，然后研究这些主题与标注目的的相关性。音色被理解为依赖于频谱和时间相关的特征，这些特征可以用对感知和声学都有意义的方式来表示。这是通过采用 Lasse Thoresen 的光谱形态分析的起点来实现的。最初为基于感知的分析而开发的符号适用于混合频谱人员系统，以指示音色的频谱质量。为了测试该系统，它被用来转录三部经典电声音乐作品的片段。除了能够比较使用同一系统转录的三个摘录的好处之外，还有一个好处是视觉表示是基于音乐中的光谱可测量质量。该乐谱系统的直观性也在听力测试中得到了探索，表明可以理解放置在五线谱系统上的频谱乐谱符号，特别是对于具有两个而不是一个声音对象的示例。