Music Navigation with Symbols and Layers: Toward Content Browsing with IEEE 1599 XML Encoding

The first book–length explanation of applications for the newIEEE Standard 1599

IEEE Standard 1599 codes music in XML such that it is renderedin symbols readable by both humans and machines. It is asignificant breakthrough over formats such as MP3s (which are onlymachine–readable) and is a valuable method of preserving and addingworth to older musical manuscripts and printed scores, because itretains the original written sheet music as a permanent image scanwhile at the same time allowing for multimedia layers (audio, aprinted copy of the score, video of a performance, alternatetakes/versions, etc.).

Music Navigation with Symbols and Layers is the firstbook–length explanation of applications for the new IEEE Standard1599, IEEE Recommended Practice for Defining a CommonlyAcceptable Musical Application Using XML. Written and edited bythe creators of the standard and geared towards professionalsworking in all aspects of the electronic and computer musicindustry this book:

Presents a manual that teaches how to encode music with IEEE1599 Considers a wide and genuinely global span of music, includingclassical, jazz, pop, opera, and even non–Western music such asAfrican drum bands or Indian melodic improvisations Posits that applications built with IEEE 1599 can move musicfrom something we listen to to something we interact with and trulyunderstand Runs the gamut from commercial (iPods and other portabledevices) to educational (tutorials on music theory) to historicaland cultural (music preservation)

Music Navigation with Symbols and Layers is a must–havereference for anyone wanting to create a computer application usingIEEE 1599.

Preface xi

A Brief Introduction to the IEEE 1599 Standard xv
Denis L. Baggi and Goffredo M. Haus

List of Contributors xvii

1 THE IEEE 1599 STANDARD 1
Denis L. Baggi and Goffredo M. Haus

1.1 Introduction 1

Important Features of IEEE 1599 2

Examples of Applications of IEEE 1599 to Increase MusicEnjoyment 3

Example I: A Score with Different Versions: King PorterStomp, by Jelly Roll Morton 6

Example II: A Jazz Piece with No Score: CrazyRhythm 6

Example III: An Opera Using the Composer s Manuscript:Tosca, by Giacomo Puccini 9

Example IV: Peaches en Regalia, by Frank Zappa9

Example V: Il mio ben quando verrà, byGiovanni Paisiello 12

Example VI: Brandenburg Concerto No. 3, by J.S. Bach 14

Example VII: Blues, a Didactical Tool to Learn JazzImprovisation 14

Example VIII: La caccia, from AntonioVivaldi s Four Seasons ( Autumn ) 16

Example IX: A Musicological Fantasy: Tauhid, a Piece of FreeJazz 17

Conclusions 19

Acknowledgments 19

References 19

2 ENCODING MUSIC INFORMATION 21
Luca A. Ludovico

2.1 Introduction 21

2.2 Heterogeneous Descriptions of Music 22

2.3 Available File Formats 23

2.4 Key Features of IEEE 1599 24

2.5 Multi–Layer Structure 25

2.6 The Logic Layer 27

2.7 The Spine 29

2.7.1 Inter–layer and Intra–layer Synchronization 31

2.7.2 Virtual Timing and Position of Events 32

2.7.3 How to Build the Spine 33

References 36

3 STRUCTURING MUSIC INFORMATION 37
Adriano Baratè and Goffredo M. Haus

3.1 Introduction 37

3.2 Music Objects and Music Algorithms 38

3.2.1 Music Objects 38

3.2.2 Music Algorithms 38

3.2.3 Music Objects and Music Algorithms in IEEE 1599 39

3.3 Petri Nets 39

3.3.1 Petri Nets Extension: Hierarchy 40

3.3.2 Petri Nets Extension: Probabilistic Arc Weights 43

3.4 Music Petri Nets 44

3.4.1 Music Petri Nets in IEEE 1599 47

3.5 Music Analysis Using Music Petri Nets 47

3.6 Real–Time Interaction with Music Petri Nets 50

3.7 Conclusions 55

References 55

4 MODELING AND SEARCHING MUSIC COLLECTIONS 57
Alberto Pinto

4.1 Introduction 57

4.2 Describing Music Content 58

4.2.1 Music Search Engines 59

4.3 Music Description in IEEE 1599 60

4.3.1 Chord Grid Objects 64

4.3.2 Petri Net Objects 65

4.3.3 Analysis Objects 65

4.3.4 MIR Objects 66

4.4 The Theoretical Framework 66

4.4.1 The Model Perspective 66

4.4.2 Categories 67

4.5 Music Modeling and Retrieval in IEEE 1599 67

4.5.1 MIR Model 68

4.5.2 MIR Object 69

4.5.3 MIR Subobject 70

4.5.4 MIR Morphisms 70

4.5.5 MIR Features 70

4.5.6 GraphXML Encoding 71

4.6 Case Study: Graph–Categorial Modeling 72

4.6.1 Content Description 72

4.6.2 Content Retrieval 72

4.6.3 MIR Model 73

4.6.4 MIR Object and Subobject 74

4.6.5 MIR Morphism 75

References 75

5 FEATURE EXTRACTION AND SYNCHRONIZATION AMONG LAYERS77
Antonello D Aguanno, Goffredo M. Haus, and Davide A.Mauro

5.1 Introduction 77

5.2 Encoding Synchronization Information 78

5.2.1 Extraction of Synchronization Data 82

5.2.2 Case Study 84

5.3 Overview of Synchronization Algorithms 84

5.4 VarSi: An Automatic Score–to–Audio Synchronization AlgorithmBased on the IEEE 1599 Format 88

5.4.1 Score Analysis 89

5.4.2 Audio Analysis 90

5.4.3 Decisional Phase 91

References 94

6 IEEE 1599 AND SOUND SYNTHESIS 97
Luca A. Ludovico

6.1 Introduction 97

6.2 From Music Symbols to Sound Synthesis 98

6.2.1 Translating Symbols into a Performance Language 99

6.2.2 Interpretative Models 105

6.2.3 Audio Rendering and Synchronization 106

6.3 From Sound Synthesis to Music Symbols 108

6.4 An Example of Encoding 110

6.5 Conclusions 113

References 114

7 IEEE 1599 APPLICATIONS FOR ENTERTAINMENT AND EDUCATION115
Adriano Baratè and Luca A. Ludovico

7.1 Introduction 115

7.2 IEEE 1599 for Entertainment 116

7.3 IEEE 1599 for Music Education 117

7.4 IEEE 1599–Based Music Viewers 118

7.5 Case Studies 120

7.5.1 Navigating and Interacting with Music Notation and Audio120

7.5.2 Musicological Analysis 121

7.5.3 Instrumental and Ear Training 126

7.5.4 IEEE 1599 Beyond Music 132

References 132

8 PAST PROJECTS USING SYMBOLS FOR MUSIC 133
Denis L. Baggi

8.1 Brief History 133

8.2 Bass Computerized Harmonization (BA–C–H) 134

8.3 Harmony Machine 135

8.4 NeurSwing, an Automatic Jazz Rhythm Section Built withNeural Nets 141

8.5 The Paul Glass System 145

8.6 A Program That Finds Notes and Type of a Chord and Plays It147

8.7 Summary of Projects 149

8.8 Conclusions 150

References 150

Appendix A. Brief History of IEEE 1599 Standard, andAcknowledgments 151

Appendix B. IEEE Document–Type Defi nitions (DTDs) 153

Appendix C. IEEE 1599 Demonstration Videos 177

Index 179

DENIS L. BAGGI was a faculty member at the University ofApplied Sciences (SUPSI) near Lugano in Southern Switzerland andthe director of the Laboratory for Semantic Systems and Multimedia.He was the chairman of the IEEE–SA Working Group in charge ofStandard IEEE 1599 to encode and represent music, and is presentlythe CEO of Think–Lab.ch, a Swiss company dedicated toinnovation.

GOFFREDO M. HAUS is the Director of the Department ofComputer Science and Dean of the Department of Information andCommunication Technology at the University of Milan. He haspublished nearly 100 scientific papers, dozens of other kinds ofpublications, and several books and CD–ROMs concerning computerapplications in music.

Both Baggi and Haus cofounded the IEEE Computer Society TaskForce on Computer Generated Music.