Od syntezatora dźwięku do symulatora wykonawcy i słuchacza robota

Apply

Od syntezatora dźwięku do symulatora wykonawcy i słuchacza robota

Journal Title: Teoria Muzyki. Studia, Interpretacje, Dokumentacje - Year 2017, Vol 10, Issue 10

Abstract

One thing never changes. It is the fact, that everything changes. We shall remember this, as we may be approaching the verge of an important change in the history of music and art. A change that might alter our place not only in the process of creation, but also a place in art itself. This change will be caused by the very tools we now so eagerly embrace and develop – computers and computer algorithms. At the end we may find ourselves either on a much higher level of possibilities, with powerful tools that understand us and free us from laborious tasks while we deal with ideas, imagination, and meaning, or we may be left out to become passive observers, with computers performing everything themselves, from creation to reception. A computer – creator, a computer – performer, and a computer – critic, they are all here, or very close. First there was an instrument. Inventions in the field of electroacoustics allowed us to build electronic instruments. Many of them had a quality unattainable for traditional instruments – their sound could be controlled to a much larger degree and with more precision. Soon, as synthesizers, they were simulating acoustic instruments, and were used instead. Then there was a performer. Unlike acoustic instruments, their electronic counterparts were controlled with signals that did not have to originate from a human player. Keyboards and other man-controlled devices have been replaced with sequencers. Initially, they replayed simple sequences, too regular and plain to be mistaken for a real performance. But a research on human performance allowed to measure small variations that occur in different musical contexts and to recreate these in a form of algorithms. While still not perfect, such advanced sequencers can be considered a substitute for real performers. But can a computer ultimately replace an auditor? We have to consider such possibility. This process starts in large repositories of digital music recordings which require an efficient method to automatically evaluate and categorise their ever expanding contents. Computer programmes can utilise signal processing and statistical methods enhanced with elements of artificial intelligence to analyse the recordings and automatically assign them with a group of tags. Consequently, tags can be used to select music presented to a particular human listener, depending on a user profile. If a programme evaluates and decides what is to be presented, then does it not result in the fact that such programme might become a computer critic? Even if a computer can be a critic, is it possible for it to become a creator? “Deep learning” seems to be the answer. It is a set of methods for modelling high level abstractions, and it is already utilised to create music without human interference. An important question is: will it become a valuable support for a creator of art, or will it replace the creator? And if the latter becomes true, what will it base its learning process on?

Authors and Affiliations

Marek Pluta, Rafał Jacek Delekta

Keywords

EP ID EP329455
DOI 10.26377/22998454tm.17.10.015
Views 121
Downloads 0