MIDIToolbox操作指南.pdf-资料库

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第1页.png

第1页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第2页.png

第2页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第3页.png

第3页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第4页.png

第4页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第5页.png

第5页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第6页.png

第6页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第7页.png

第7页 / 共95页

02423a53-2319-4b7a-ac10-62f1ecf166d8.pdf-第8页.png

第8页 / 共95页

toolbox F5# E5 D5 C5 A4# G4# F4# E4 h c t i P D4 0 1 2 3 6 7 8 4 5 Time in seconds MATLAB Tools for Music Research Tuomas Eerola & Petri Toiviainen

MIDI Toolbox MATLAB Tools for Music Research Tuomas Eerola & Petri Toiviainen Music Cognition Group

Copyright ©: Tuomas Eerola & Petri Toiviainen Cover & layout: Tuomas Eerola & Petri Toiviainen Publisher: Department of Music, University of Jyväskylä, Finland Printing and binding: Kopijyvä, Jyväskylä, Finland ISBN: 951-39-1796-7 (printed version) ISBN: 951-39-1795-9 (pdf version) Document data: Eerola, T. & Toiviainen, P. (2004). MIDI Toolbox: MATLAB Tools for Music Research. University of Jyväskylä: Kopijyvä, Jyväskylä, Finland. Electronic version available from: http://www.jyu.fi/musica/miditoolbox/

4 ■ CONTENTS CONTENTS ■ CHAPTER 1 – INTRODUCTION.................................................................................... 6 ■ CHAPTER 2 – INSTALLATION..................................................................................... 8 ■ CHAPTER 3 – BASIC OPERATIONS............................................................................ 9 ■ CHAPTER 4 – EXAMPLES ........................................................................................... 15 ■ EXAMPLE 1: VISUALIZING MIDI DATA............................................................... 15 ■ EXAMPLE 2: MELODIC CONTOUR........................................................................ 21 ■ EXAMPLE 3: KEY-FINDING................................................................................... 24 ■ EXAMPLE 4: METER-FINDING.............................................................................. 29 ■ EXAMPLE 5: MELODIC SEGMENTATION.............................................................. 32 ■ EXAMPLE 6: MELODIC EXPECTATIONS............................................................... 34 ■ EXAMPLE 7: MELODIC COMPLEXITY.................................................................. 37 ■ EXAMPLE 8: ANALYZING MIDI COLLECTIONS .................................................. 38 ■ EXAMPLE 9: MELODIC SIMILARITY..................................................................... 41 ■ EXAMPLE 10: CREATING SEQUENCES ................................................................. 43 ■ REFERENCES.......................................................................................................... 45 ■ CHAPTER 5 – FUNCTION REFERENCE................................................................... 48 ■ ALPHABETICAL INDEX OF FUNCTIONS................................................................ 96 ■ MIDI Toolbox ■

■ CHAPTER 1 – INTRODUCTION 6 CHAPTER 1 – INTRODUCTION MIDI Toolbox provides a set of Matlab functions, which together have all the necessary machinery to analyze and visualize MIDI data. The development of the Toolbox has been part of ongoing research involved in topics relating to musical data- mining, modelling music perception and decomposing the data for and from perceptual experiments. Although MIDI data is not necessarily a good representation of music in general, it suffices for many research questions dealing with concepts such as melodic contour, tonality and pulse finding. These concepts are intriguing from the point of view of music perception and the chosen representation greatly affects the way these issues can be approached. MIDI is not able to handle the timbre of music and therefore it unsuitable representation for a number of research questions (for summary, see Hewlett and Selfridge-Field, 1993-94, p. 11-28). All musical signals may be processed from acoustic representation and there are suitable tools available for these purposes (e.g. IPEM toolbox, Leman et al., 2000). However, there is a body of essential questions of music cognition that benefit from a MIDI-based approach. MIDI does not contain notational information, such as phrase and bar markings, and neither is that information conveyed in explicit terms to the ears of music listeners. Consequently, models of music cognition must infer these musical cues from the pitch, timing and velocity information that MIDI provides. Another advantage of the MIDI format is that it is extremely wide-spread among the research community as well as having a wider group of users amongst the music professionals, artists and amateur musicians. MIDI is a common file format between many notation, sequencing and performance programs across a variety of operating systems. Numerous pieces of hardware exist that collect data from musical performances, either directly from the instrument (e.g. digital pianos and other MIDI instruments) or from the movements of the artists (e.g. motion tracking of musician’s gestures, hand movements etc.). The vast majority of this technology is based on MIDI representation. However, the analysis of the MIDI data is often developed from scratch for each research question. The aim of MIDI Toolbox is to provide the core representation and functions that are needed most often. These basic tools are designed to be modular to allow easy further development and tailoring for specific analysis needs. Another aim is to facilitate efficient use and to lower the “threshold of practicalities”. For example, the Toolbox can be used as teaching aid in music cognition courses. This documentation provides a description of the Toolbox (Chapter 1), installation and system requirements (Chapter 2). Basic issues are explained in Chapter 3. Chapter 4 demonstrates the Toolbox functions using various examples. The User’s Guide does not describe any of the underlying theories in detail. Chapter 5 focuses on a collection format and Chapter 6 is the reference section, describing all functions in the Toolbox. The online reference documentation provides direct hypertext links to specific Toolbox functions. This is available at http://www.jyu.fi/musica/miditoolbox/ ■ MIDI Toolbox ■

CHAPTER 1 – INTRODUCTION ■ 7 This User’s Guide assumes that the readers are familiar with Matlab. At the moment, the MIDI Toolbox is a collection of Matlab functions that do not require any extra toolboxes to run. Signal processing and Statistics toolboxes – both available separately from Mathworks – offer useful extra tools for the analysis of perceptual experiments. MIDI Toolbox comes with no warranty. This is free software, and you are welcome to redistribute it under certain conditions. See License.txt for details of GNU General Public License. We would like to thank various people contributing to the toolbox. The conversion to and from MIDI file is based on the C source code by Piet van Oostrum, which, in turn, uses the midifile library written by Tim Thompson and updated by Michael Czeiszperger. Brian Cameron found out some sneaky bugs in the aforementioned C source code. Micah Bregman helped to check parts of the manual and wrote out some new functions. Comments, suggestions or questions? Many functions are still not completely tested in MIDI Toolbox version 1.0. Check the online forum for corrections and revisions: http://www.jyu.fi/musica/miditoolbox/forum.html Alternatively, you can report any bugs or problems to: Tuomas Eerola, Petri Toiviainen {ptee, ptoiviai}@cc.jyu.fi Department of Music University of Jyväskylä P.O. BOX 35 40014 University of Jyväskylä Finland ■ MIDI Toolbox ■

■ CHAPTER 1 – INTRODUCTION 8 CHAPTER 2 – INSTALLATION Availability The whole toolbox is available either as a zipped package from the internet (http://www.jyu.fi/musica/miditoolbox/). Installation Unpack the MIDI Toolbox file package you have downloaded. For this, use a program like Winzip for Windows and Stuffit Expander for Macintosh. This will create a directory called miditoolbox. Secondly, a version of the Matlab program needs to be installed (see www.mathworks.com). Thirdly, the Toolbox needs to be defined in the Matlab path variable. Windows (98, 2000, XP) The MIDI Toolbox version 1.0 is compatible with Matlab 5.3 and Matlab 6.5. Macintosh (OS X) The MIDI Toolbox version 1.0 is compatible with Matlab 6.5 for Macintosh. Linux Currently not tested but should be compatible. ■ MIDI Toolbox ■

9 CHAPTER 3 – BASIC OPERATIONS ■ CHAPTER 3 – BASIC OPERATIONS good tutorials exist in the Internet, syntax. Many http://www.math.ufl.edu/help/matlab-tutorial/ http://www.math.mtu.edu/~msgocken/intro/intro.html http://www.helsinki.fi/~mjlaine/matlab/index.html (in Finnish) http://www.csc.fi/oppaat/matlab/matlabohje.pdf (in Finnish) Basic issues In this tutorial, we assume that the reader has basic knowledge of the Matlab command see: In the following examples, the commands that are typed to Matlab command prompt are written in monospaced font and are preceded by the » sign. Help is also available within the Matlab session. For example, to understand what a particular function does, type help and the name of the function at the command prompt. For example, to obtain information about how the pitch-class distribution function works, type: To see a list of all available functions in the Toolbox, type: » help miditoolbox » help pcdist1 Reading MIDI files into Matlab The basic functions in MIDI Toolbox read and manipulate type 0 and type 1 MIDI files. The following command reads and parses a MIDI file called laksin.mid and stores it as a matrix of notes called nmat in Matlab’s workspace: This particular MIDI file contains the first two verses of a Finnish Folk song called "Läksin minä kesäyönä" (trad.). » nmat = readmidi('laksin.mid'); Basic terms Notematrix (or nmat) refers to a matrix representation of note events in a MIDI file. We can now type nmat and see what the notematrix of the folk song looks like. ■ MIDI Toolbox ■

资料库

MIDIToolbox操作指南.pdf

相关推荐

课程资源

热门标签

最新资料