Digital music

Music is traditionally retrieved by title, composer or subject classification. It is possible, with current technology, to retrieve music from a database on the basis of a few notes sung or hummed into a microphone. This paper describes the implementation of such a system, and discusses several issues pertaining to music retrieval. We first describe an interface that transcribes acoustic input into standard music notation. We then analyze string matching requirements for ranked retrieval of music and present the results of an experiment which tests how accurately people sing well known melodies. The performance of several string matching criteria are analyzed using two folk song databases. Finally, we describe a prototype system which has been developed for retrieval of tunes from acoustic input.

INTRODUCTION Musical performance with entirely new types of computer instruments is now commonplace, as a result of the availability of inexpensive computing hardware, of new sensors for measuring physical parameters such as force and position, and of new software for ...

The increasing pervasiveness of the internet, broadband connections and the emergence of digital compression technologies have dramatically changed the face of digital music piracy. Digitally compressed music files are essentially a perfect public economic good, and illegal copying of these files has increasingly become rampant.

W e study the offers submitted by consumers to a large Name-Your-Own-Price (NYOP) online retailer. A distinctive feature of this retailer is that it allows consumers to repeatedly submit offers on one and the same product. While consumers could identify the threshold price (the minimum price for which the retailer is willing to sell) by incrementing their offer in small steps in each consecutive round, such a strategy would require them to go through many additional online transactions. We define frictional cost as the disutility that the consumer experiences when conducting an online transaction, such as submitting an offer. Thus, in our setting, consumers trade off a direct financial value (lower price) for frictional costs. Based on a consumer choice model capturing this trade-off, we use the observed consumer behavior to reconstruct the frictional cost parameters for the consumers in our sample. We show that, perhaps contrary to the general wisdom, frictional costs in electronic markets are substantial, with median values ranging from EUR 3.54 for a portable digital music player (MP3) to EUR 6.08 for a personal digital assistant (PDA). We find that consumers who have gathered experience with the NYOP channel in previous transactions exhibit lower frictional costs than consumers who use the channel for the first time. Surprisingly, sociodemographic variables do not help to explain the variation in frictional costs.

Active learning is a subfield of machine learning based on the idea that the accuracy of an algorithm can be improved with fewer training samples if it is allowed to choose the data from which it learns. We present the results for Support Vector Machine (SVM) active learning experiments for music mood tagging based on a multi-sample selection strategy that chooses samples according to their proximity to the boundary, their proximity to points in the training set and the density around them. The influence of those key active learning parameters is assessed by means of ANalysis Of Variance (ANOVA). Using these analyses we demonstrate the efficiency of active learning compared to typical fulldataset batch learning: our method allows to tag music by mood more efficiently than a regular approach, requiring fewer instances to obtain the same performance than using random sample selection methods.

We present a framework for analyzing the structure of digital media streams. Though our methods work for video, text, and audio, we concentrate on detecting the structure of digital music files. In the first step, spectral data is used to construct a similarity matrix calculated from inter-frame spectral similarity. The digital audio can be robustly segmented by correlating a kernel along the diagonal of the similarity matrix. Once segmented, spectral statistics of each segment are computed. In the second step, segments are clustered based on the selfsimilarity of their statistics. This reveals the structure of the digital music in a set of segment boundaries and labels. Finally, the music can be summarized by selecting clusters with repeated segments throughout the piece. The summaries can be customized for various applications based on the structure of the original music.

We present a novel, innovative user interface to music repositories. Given an arbitrary collection of digital music files, our system creates a virtual landscape which allows the user to freely navigate in this collection. This is accomplished by automatically extracting features from the audio signal and training a Self-Organizing Map (SOM) on them to form clusters of similar sounding pieces of music. Subsequently, a Smoothed Data Histogram (SDH) is calculated on the SOM and interpreted as a three-dimensional height profile. This height profile is visualized as a three-dimensional island landscape containing the pieces of music. While moving through the terrain, the closest sounds with respect to the listener's current position can be heard. This is realized by anisotropic auralization using a 5.1 surround sound model. Additionally, we incorporate knowledge extracted automatically from the web to enrich the landscape with semantic information. More precisely, we display words and related images that describe the heard music on the landscape to support the exploration.

Digital libraries of music have the potential to capture popular imagination in ways that more scholarly libraries cannot. We are working towards a comprehensive digital library of musical material, including popular music. We have developed new ways of collecting musical material, accessing it through searching and browsing, and presenting the results to the user. We work with different representations of music: facsimile images of scores, the internal representation of a music editing program, page images typeset by a music editor, MIDI files, audio files representing sung user input, and textual metadata such as title, composer and arranger, and lyrics.

Current research on music information retrieval and music digital libraries focuses on providing access to huge, public music collections. In this paper we consider a different, but related, problem: supporting an individual in maintaining and using a personal music collection. We analyze organization and access techniques used to manage personal music collections (primarily CDs and MP3 files), and from these behaviors, to suggest user behaviors that should be supported in a personal music digital library (that is, a digital library of an individual's personal music collection). Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page.

One major task of a digital music library (DML) is to provide techniques to locate a queried musical pattern in all pieces of music in the database containing that pattern. For a survey of several computational tasks related to this kind of data retrieval we refer to Crawford et al. [3]. Existing DMLs like MELDEX [1], Themefinder [4], and the Sonoda-Muraoka-System [7] work with melody databases relying on score-like information. Retrieval and matching are performed in a fault-tolerant way by string-based methods which mainly take into account pitch information. Generally, rhythm plays only a subordinate role. The music dictionary of Barlow and Morgenstern [2] shows that music retrieval based on pitch information only leads to results with typically too many false matches. (An example of such absurd matches is given in Selfridge-Field [6], p. 27.) We are convinced that both pitch and rhythm are crucial for recognizing melodies. In the more general context of polyphonic music, one is even forced to consider pitch and rhythm information. PROMS, a web-based computer-music service under development at the University of Bonn, Germany, is part of the MiDiLiB project [5]. The aim of PROMS is to design and to implement PROcedures for Music Search. Our discussion will take place in a rather general setting: we assume that our database contains several kinds of music such as polyphonic and homophonic music as well as melodies. We also use score-like information. A query to the database is a fragment of a piece of music. This could be a melody or a certain figure of an accompaniment. The task is now to locate efficiently all occurences of this fragment in all pieces of music in the database. We have designed and implemented a web-based computer-music service that enables seaching in polyphonic music. In contrast to the above mentioned systems, PROMS is not string-based but set-oriented. The basic objects within the PROMS system are single notes, specified by its onset time t, its pitch p, and its duration d. A piece of music is then a finite set M of notes. Our database consists of a sequence of N pieces of music in the MIDI format: D 1 ,. .. , D N. Similarly, a query is itself a finite set Q of notes. Thus there is no principal difference between a piece D i of music in the database and a query Q. However, typically, D i is much larger than Q. An occurrence is a pair * This work was supported in part by Deutsche Forschungsgemeinschaft grants CL 64/3-1 and CL 64/3-4.

This paper describes the adaptation of an existing model of hu- man information processing for the categorization of digital musi- cal instruments in terms of performance context and behavior. It further presents a visualization intended to aid the analysis of ex- isting DMIs and the design of new devices. Three new interfaces constructed by the authors are examined within this

Digital products have the property that they can be copied almost costlessly. This makes them candidates for non-commercial copying by final consumers. Because the copy of a copy typically does not deteriorate in quality, copying products can become a wide-spread phenomenon - this can be illustrated by the surge of file-sharing networks. In this paper we provide a critical overview

This paper describes the T-Stick, a new family of digital musical instruments. It presents the motivation behind the project, hardware and software design, and presents insights gained through collaboration with performers who have collectively practised and performed with the T-Stick for hundreds of hours, and with composers who have written pieces for the instrument in the context of McGill University's Digital Orchestra project. Each of the T-Sticks is based on the same general structure and sensing platform, but each also differs from its siblings in size, weight, timbre and range.

We present a system capable of performing similarity queries against a large archive of digital music. Users are able to search for songs which "sound similar" to a given query song, thereby aiding the navigation and discovery of new music in such an archive. Our technique is based on reduction of the music data to a feature space of relatively small dimensionality (1248 feature dimensions per song); this is accomplished using a set of feature extractors which derive frequency, amplitude, and tempo data from the encoded music data. Queries are then performed using a k-nearest neighbor search in the feature space. Our system allows subsets of the feature space to be selected on a per-query basis.

In this paper, a new system for the automatic transcription of singing sequences into a sequence of pitch and duration pairs is presented. Although such a system may have a wider range of applications, it was mainly developed to become the acoustic module of a queryby-humming (QBH) system for retrieving pieces of music from a digitized musical library. The first part of the paper is devoted to the systematic evaluation of a variety of state-of-the art transcription systems. The main result of this evaluation is that there is clearly a need for more accurate systems. Especially the segmentation was experienced as being too error prone ( ¾¼ % segmentation errors). In the second part of the paper, a new auditory model based transcription system is proposed and evaluated. The results of that evaluation are very promising. Segmentation errors vary between 0 and 7 %, dependent on the amount of lyrics that is used by the singer. The paper ends with the description of an experimental study that was issued to demonstrate that the accuracy of the newly proposed transcription system is not very sensitive to the choice of the free parameters, at least as long as they remain in the vicinity of the values one could forecast on the basis of their meaning.

The Clavinet is an electromechanical musical instrument produced in the mid-twentieth century. As is the case for other vintage instruments, it is subject to aging and requires great effort to be maintained or restored. This paper reports analyses conducted on a Hohner Clavinet D6 and proposes a computational model to faithfully reproduce the Clavinet sound in real time, from tone generation to the emulation of the electronic components. The string excitation signal model is physically inspired and represents a cheap solution in terms of both computational resources and especially memory requirements (compared, e.g., to sample playback systems). Pickups and amplifier models have been implemented which enhance the natural character of the sound with respect to previous work. A model has been implemented on a real-time software platform, Pure Data, capable of a 10-voice polyphony with low latency on an embedded device. Finally, subjective listening tests conducted using the current model are compared to previous tests showing slightly improved results.

Emerging web services technologies provide an open infrastructure for automated business interaction, thereby creating new opportunities for business actors to collaborate within a networked constellation of enterprises via the Internet. The basis for a viable network of web services (the supporting information system of such a networked constellation of enterprises) is a value model that shows sound value propositions to all actors involved. Requirements engineering techniques can be developed to support: (1) exploring alternative business models, and (2) evaluating alternatives on their economic viability, leading into the design and implementation of technical systems. In this paper we present a Business-oriented Approach Supporting web Services Idea Exploration (BASSIE), which exploits the synergy between the agent-and goal-oriented i* framework and the value-based e 3 value framework. The approach iterates between exploration of structural alternatives and qualitative evaluation using i*, and quantitative modeling and evaluation of business value using e 3 value. The approach is illustrated with a real-life case study in digital music distribution.

digital music, MP3, affordances, new media, tangibility A number of new technologies such as the MP3 music compression format and the Napster file sharing network are creating new opportunities for consumers' activities with music media. In this paper we explore current practice with both conventional and new forms of media in order to better understand what the future may hold for music consumers. We interviewed 36 music enthusiasts: 12 teenaged and 12 adult conventional music users, and 12 early adopters of MP3 technologies. The results show that rather than new music technologies replacing the old, they supplement current formats by allowing users to exploit a different set of affordances. We discuss these uses drawing implications for new music devices, services and software.

New digital musical instruments designed for professional and trained musicians can be quite complex and challenging, offering as a counterpart a great amount of creative freedom and control possibilities to their players. On the other hand, instruments designed for amateur musicians or for audiences in interactive sound installations, tend to be quite simple, often trying to bring the illusion of control and interaction to their users, while still producing 'satisfactory' outputs. Logically, these two classes of instruments are often mutually exclusive. But wouldn't it be possible to design instruments that can appeal to both sectors? In this paper we will show , with two projects developed in our research group, how visual feedback can highly increase the intuitiveness of an interactive music system, making complex principles understandable.

This paper explores a longitudinal approach to the qualitative evaluation of a set of digital musical instruments, which were developed with a focus on creativity and exploration. The instruments were lent to three electronic musicians/composers for a duration of four weeks. Free exploration periods and a compositional task evaluated using semi-structured show and tell sessions revealed that the context of use had a major influence on how the instruments were experienced by the test persons. Central issues revolved around the learning/ adoption process, and around the importance of playability, explorability, and connectivity.

The explosion of digital music has created in the recent years an urgent need for powerful knowledge management techniques and tools. Without such tools, users are confronted to huge music catalogues they cannot fully exploit. The very nature of music calls for the development of specific knowledge management techniques: on the one hand, the goals of users are ill-defined, or

The goal of this paper is to sketch the value chain and the business models of the online distribution of music. The perspective of the online digital music market is rather deceiving but the opportunities seem to remain high. Considering the rise of the P2P networks of free music digital files, it seems reasonable to assume this new way to distribute music meets consumer needs. Based on a review of the literature and executive interviews, the paper presents the traditional distribution models; then it addresses how P2P piracy deals with the copyright issues, and describes some emergent business models that could be an answer to illegal digital music distribution.

Heart rate data reflects various physiological states such as biological workload, stress at work and concentration on tasks, drowsiness and the active state of the autonomic nervous system. This paper proposes system to indicate heart rate using musical data. Music changes physiological states for the better by relaxing people, or contributing to patient treatment through music therapy. Information in the form of music is advantageous because it does not hinder work as does verbal information and it contains more information than warning noises.

In this paper, after giving some possible definitions for expressiveness, we examine the problem of expressiveness in digital musical instruments, which tends to involve using specific gestures to obtain an expressive sound rather than performing expressive gestures. Some of the particular features of digital musical instruments, such as pitch control, dynamic control and the possibility of exploring sound palettes, are described and some practical examples given. Lastly, several musical implications of the gestures used to obtain musical expressiveness are discussed, from the pedagogical and other related points of view.

This paper focuses on the theoretic issues of the collaborative project Canções dos Olhos (Paulo C. Chagas -composition, Johannes Birringer -choreography and video, Veronica Endo -dance) in the context of Interaktionslabor, a laboratory for interactive media, sound, design, digital video, telecommunications and performance, taking place on the site of the former coal mine Göttelborn, Germany. The work, inspired by the novel "Blindness" by José Saramago, explores the cognitive and aesthetic dimensions of blindness in terms of embodiment experience. The project was conceived as an "intermedia" song cycle resulting in a DVD and an audiovisual installation.

To some extent listening to digital music via storage devices has led to a loss of part of the physical experience associated with earlier media formats such as CDs and LPs. For example, one could consider the role of album covers in music appreciation. Previous efforts at making music interaction more tangible have focused mainly on access issues. A case study is presented in which several content attributes of Mp3 formatted music as well as control access are made more visible and tangible. Play lists, music rhythm, volume, and navigational feedback were communicated via multicolored light displayed in a tangible interface. Users were able to physically interact with music collections via the MusicCube, a wireless cube-like object, using gestures to shuffle music and a rotary dial with a button for song navigation and volume control. Speech and non-speech feedback were given to communicate current mode and song title. The working prototype was compared to an Apple iPod, along the dimensions of trust, engagement, ergonomic and hedonic qualities, and appeal. Subjects rated the MusicCube higher on scales associated with hedonic qualities, while the iPod was preferred for ergonomic qualities. Results on trust measures were found to correlate with ergonomic qualities, while sense of engagement related to hedonic aspects. Subjects generally valued the expressive and tangible interaction with music collections. Next design steps will focus on increasing ergonomic aspects of the MusicCube while maintaining a high hedonic rating.

This paper describes the design and implementation of a framework designed to aid collaborative development of a digital musical instrument mapping layer 1 . The goal was to create a system that allows mapping between controller and sound parameters without requiring a high level of technical knowledge, and which needs minimal manual intervention for tasks such as configuring the network and assigning identifiers to devices. Ease of implementation was also considered, to encourage future developers of devices to adopt a compatible protocol.

The widespread availability of digitized music collections and mobile music players have enabled us to listen to music during many of our daily activities, such as physical exercise, commuting, relaxation, and many people enjoy this. A practical problem that comes along with the wish to listen to music is that of music retrieval, the selection of desired music from a music collection. In this paper, we propose a new approach to facilitate music retrieval. Modern smart phones are commonly used as music players and are already equipped with inertial sensors that are suitable for obtaining motion information. In the proposed approach, emotion is derived automatically from arm gestures and is used to query a music collection. We derive predictive models for valence and arousal from empirical data, gathered in an experimental setup where inertial data recorded from arm movements are coupled to musical emotion. Part of the experiment is a preliminary study confirming that human subjects are generally capable of recognizing affect from arm gestures. Model validation in the main study confirmed the predictive capabilities of the models.

Popular acceptance of the mp3 digital music standard has greatly increased the complexity of organizing and playing large music collections. Existing digital music systems do not adequately support exploration of a collection, nor do they cater to multi-user interaction in a social setting. In this paper, we present the design of a ubiquitous system that utilizes spatial visualization to support exploration and social interaction with a large music collection. Our interface is based on the interaction semantic of influence, which allows users to affect and control the mood of music being played without the need to select a set of specific songs. This design is inspired, to some extent, by Gaver's work on ludic design. We implemented a prototype as a proof of concept of our design. User testing demonstrates that our system encourages participation and strengthens social cohesion. Our work contributes to interactive interface research in that it extends the utility of map-based visualization of digital music.

, as well as in a handful of related examples, we argue that one can indeed identify specific aural qualities associated with digital sound, and that these qualities may be used to achieve different aesthetic effects as well as to shed light on mediation and medium specificity as such.

There has been a recent explosion of interest in digital music libraries. In particular, interactive melody retrieval is a striking example of a search paradigm that differs radically from the standard full-text search. Many different techniques have been proposed for melody matching, but the area lacks standard databases that allow them to be compared on common grounds-and copyright issues have stymied attempts to develop such a corpus. This paper focuses on methods for evaluating different symbolic music matching strategies, and describes a series of experiments that compare and contrast results obtained using three dominant paradigms.

In this paper, a subbands multirate architecture is presented for audio signal recovery. Audio signal recovery is a common problem in digital music signal restoration field, because of corrupted samples that must be replaced. The subband approach allows for the reconstruction of a long audio data sequence from forward-backward predicted samples. In order to improve prediction performances, neural networks with spline flexible activation function are used as narrow subband nonlinear forward-backward predictors. Previous neural-networks approaches involved a long training process. Due to the small networks needed for each subband and to the spline adaptive activation functions that speed-up the convergence time and improve the generalization performances, the proposed signal recovery scheme works in on-line (or in continuous learning) mode as a simple nonlinear adaptive filter. Experimental results show the mean square reconstruction error and maximum error obtained with increasing gap length, from 200 to 5000 samples for different musical genres. A subjective performances analysis is also reported. The method gives good results for the reconstruction of over 100 ms of audio signal with low audible effects in overall quality and outperforms the previous approaches.

Most of the digital musical instruments developed so far were specifically designed for music performances and the interfaces are disruptive. They are not generally available and some of them were made with expensive technologies, making them inaccessible to general users. This paper presents an attempt to make use of Wii Controller(or Wiimote), a low cost and readily available game controller developed by Nintendo to develop an interactive music performance system on PC platform.

Digital Rights Management (DRM) technology ensures protection of rights on distributed music over the Internet. Existing DRM systems emphasize the protection on Businessto-Consumer (B2C) distribution model, but pay very little attention on the protection of Consumer-to-Consumer (C2C) distribution model-referring to peer-to-peer sharing. In this paper, we propose a license management model to support not only the music distributor to consumer (B2C) music distribution, but also consumer-to-consumer music sharing. The proposed license management model consists of two types of license; namely official license and peer license. We proposed to implement these licenses using XrML.

Online Digital Music Libraries are becoming increasing com- mon and more sophisticated; however, availability of infor- mation on how users access and navigate these resources is limited. This type of information is crucial for improv- ing user interface design and for providing users with better supported services. Here we present an analysis of the logs for our digital music library,

This article explores the role visual browsing can play within a digital music library. The context to the work is provided through a review of related techniques drawn from the fields of digital libraries and human computer interaction. Implemented within the open source digital library toolkit Greenstone, a prototype system is described that combines images located through textual metadata with a visualisation technique known as collaging to provide a leisurely, undirected interaction with a music collection. Emphasis in the article is given to the augmentations of the basic technique to work in the musical domain.