Computational musicology

Computational musicology is a fun area where music and computer science meet. It uses computers and digital tools to study how music works. This field includes subjects like mathematical music theory, computer music, and music information retrieval.

People have used computers to create music since the 1950s. The serious study of music with computers began in the 1960s. Today, researchers in computational musicology explore many ways to understand music using technology. This field connects the humanities and sciences, making it a rich and exciting subject to explore.

History

Computational musicology began in the middle of the 20th century. It builds on studying music with science, math, technology, and archiving.

In the early 1960s, people started using new ways to study music with computers. They used tools like paper tape or punch cards because computers were very expensive. Big projects focused on big questions to get funding. One early project was DARMS, supported by Columbia University and the Ford Foundation from 1964 to 1976. It aimed to create a clear way to represent music.

The 1970s focused more on specific tasks. Projects like MUSTRAN at Indiana University helped analyze music, while SCORE at Stanford University was made for printing music.

The 1980s saw a shift from big computers to personal computers, which helped the field grow. Programs like Savy PC helped study lyrics, and new ways to record music like MIDI were introduced. Researchers also studied old music manuscripts to learn more about their history.

Methods

Computational musicology studies music using computers in three main ways: sheet music, symbolic data, and audio data.

Sheet music is how music is written for people to read and play. It shows symbols that musicians use to understand the music.

Symbolic data is music stored in a way that computers can understand, not for people to read. This helps computers study the music.

Audio data is recordings of actual music sounds. It looks at features like loudness, pitch, and style to understand the music.

Applications

See also: List of music software

One of the first uses of computational musicology was creating and using music databases. Computers make it easier to handle and study large amounts of music data.

Computers can also help analyze music. Programs can work with different types of music data, from written music scores to actual sound recordings. A common way to analyze music is by using MIDI data, which records details about each note.

Computers can even help create new music. Algorithms can be used to write full songs or to improvise music. Researchers are exploring how music changes over time. Computational musicology is also used to study music from different parts of the world. For example, researchers have used these methods to study patterns in Hindustani classical music, particularly within the raga structure.

Artificial neural networks help programs learn to create music.

Research

RISM's database is a very big music database. It has more than 700,000 references to music papers. Anyone can use its search tool to find songs.

The Centre for History and Analysis of Recorded Music (CHARM) made the Mazurka Project. This project has recordings, computer tools, learning materials, and other things about the history of recording that you can download.

Computational musicology in popular culture

Sometimes, discoveries from computational musicology get a lot of attention. For example, reporters from The New Yorker wrote about musicologists Nicholas Cook and Craig Sapp. They worked at a special place called CHARM at the University of London and found out that some recordings by a pianist named Joyce Hatto were not real.

To celebrate the birthday of the famous composer Johann Sebastian Bach, Google made a special drawing on its website. People could type in their own musical notes, and a smart computer program called Coconet would create a matching tune using machine learning.