笔记 Tepora：How Generative Music Works 生成音乐是如何工作的
这是一个交互式的解释何为生成音乐的网页，其中介绍展开了两种系统，即一种为机器系统，不同部件分别工作；另一种是设计规则，然后交付实时的生成，并且将生成的依据链接到数据集，事物，还有人的交互上：其涉及 随机性Randomness，机器学习等等，其中有不少的比如Steve Reich，还有Brian Eno，John Cage的采访都非常值得翻译。
https://teropa.info/loop/#/title
这个大佬的个人页面有非常多的关于网页音乐Web Audio相关的内容
System 1
A set of things working together as parts of a mechanism or an interconnecting network; a complex whole. 一组事物作为一个机制或一个相互连接的网络的组成部分一起工作；一个复杂的整体。
Steve Reich
Though I may have the pleasure of discovering musical processes and composing the musical material to run through them, once the process is set up and loaded it runs by itself.
Brian Eno
https://inmotionmagazine.com/eno1.html
The thing about pieces like this is that they are actually of almost infinite length. They simply don't reconfigure the same way again. This is music for free in a sense. The considerations that are important, then, become questions of how the system works and most important of all, what you feed into the system.
System 2
A set of principles or procedures according to which something is done; an organized scheme or method. 一套原则或程序，根据这些原则或程序来做某事；一个有组织的计划或方法。
Terry Riley
Every performer can repeat each loop for as long as they like. As long as collectively the group stays within a few loops of each other.
With this simple blueprint you get a system that generates a new version every time it's played. No two performances will ever be the same.The system is simulated here by a quintet of marimbas. Use the slider to see how they generate a different performance every time.
After working with tape loops and repetition a lot, I had this idea of a musical universe emerging that I hadn’t heard before coming together. Since then, I’ve always liked to include something in the mix that has some chance operational quality about it. I like to continue surprising myself. Terry Riley, 2014. https://www.innerviews.org/inner/riley.html
Brian Eno 的 Reflection http://www.generativemusic.com/reflection.html
You can get “Reflection“ as an app. The generative system itself is installed to your phone. It produces an infinite stream of unique music.
这一系列的Brian Eno的相关作品 http://www.generativemusic.com/index.html
方便
Generative Software system: it makes it easy to deliver to listeners, much easier than reel-to-reel tape recorders.
Randomness
随机性
One of them is their capability to make lots of random ch/ices efficiently. And randomness is an incredibly useful generative technique.
https://codepen.io/teropa/pen/PKoYXM
John Cage
john cage was a pioneer in the use of randomness in music composition.
His method ever since the 1950s was to "ask questions rather than make choices".
This is a generative system.
问一个问题，那么你所得到的就是 set up a possibility space: That of all the possible answers.
For example, "what duration should this note have?" 然后让 random process answer the question, you generate the result.
Then the answers, instead of coming from my likes and dislikes, come from chance operations, and that has the effect of opening me to possibilities that I hadn't considered. Chance-determined answers will open my mind to the world around. — john Cage 1985
John Cage 的 采访 http://www.moredarkthanshark.org/eno_int_ammusic-sum85.html
《沉默》https://book.douban.com/subject/25731757/
Algorithms 算法
software is good at is algorithms.
That is, following a clear plan, making calculations, and doing formal reasoning/
比如说Noam Chomskky 的
https://codepen.io/teropa/pen/rzQMEG
https://codepen.io/teropa/pen/mBbPEe
与机器/数据集/实时系统联结
Helsinki 的列车随机生成
But the conceptual layer is significant. The idea that you're listening to the city's nervous system changes the way you experience the music.
This was an example of data sonification: Generative musical systems riding on events or data produced by other systems.
如果是和人类进行连接 This is the realm of interactive generative systems.
Brian Eno - Trope
将composer以及listner之间的界限混淆
http://www.generativemusic.com/trope.html
一方面，作曲家放手了，但是另一方面，the composer still has the uptimate control, as a system designer, they design the generative space the listener explores.
给予玩家更多自由 These interactive generative systems can be designed to give the user varying degrees of freedom. When there's more freedom, the interaction becomes about co-creation between a musician and a generative system.
Laurie Spiegel's “Music Mouse” from 1986 is a seminal example of such a system.It contains algorithmic and random procedures for generating and timing notes. But ultimately it lets you, the musician, control the structure of the music. The result is a collaboration between you and Music Mouse.
I know that a lot of the ways that music is integrated to our lives is changing. One of the big changes is that music is becoming a process that people can participate in - rather than a bunch of fixed, finite entities called pieces that you can listen to that are the same every time - Laurie Spiegel, 1987.
Laurie Spiegel http://retiary.org/ls/index.html
Music Mouse Computer Software by Laurie Spiegel http://retiary.org/ls/programs.html
https://teropa.info/musicmouse/
机器学习
马尔科夫链
Markov chains are an old, but still effective, example of the machine learning approach. A Markov chain is a mathematical model of how events follow other events over time. Events such as notes in a melody. You can train a Markov chain by playing it some music. It learns the music you play and builds up a graph model of it.
You can then generate new music by having the Markov chain randomly walk its graph, following the connections it learned. It's a limited method because such a simple model can only capture a limited amount of information. In this case it only learns about short-term melodic progression. But it can still be useful.
https://codepen.io/teropa/pen/bRqYVj
深度学习 neural network
https://medium.com/artists-and-machine-intelligence/neural-nets-for-generating-music-f46dffac21c0
https://magenta.tensorflow.org/performance-rnn-browser
But fundamentally they work just like Markov chains: You teach the machine how music works by showing it examples. It builds up a model of the music and then uses the model to generate new music.
This is what the generative music landscape looks like to me. It is inhabited by various kinds of systems.
Making generative music is all about the process of designing those systems.
