Standard Music Notation Research Articles

Computer Sound Synthesis in 1951: The Music of CSIRAC

Computer Music Journal, 28:1, pp. 10–25, Spring 2004 q 2004 Massachusetts Institute of Technology. The Australian-built ‘‘automatic computer’’ initially known as the CSIR Mk1, and later known as CSIRAC, was one of the world’s earliest storedprogram electronic digital computers (Williams 1997). (See Figure 1.) Coincidentally, it may also have been the Ž rst computer to play music, even though later work done elsewhere in the 1950s is clearly the origin of computer music as we know the Ž eld today. Developed in Sydney in the late 1940s by the Council for ScientiŽ c and Industrial Research (CSIR), the CSIR Mk1 ran its Ž rst program in November 1949. Geoff Hill, a mathematician and Australia’s Ž rst real software engineer, programmed the CSIR Mk1 to play popular musical melodies through its loudspeaker starting in 1951, if not 1950. The CSIR Mk1 was moved to the University of Melbourne in June 1955 and renamed CSIRAC (McCann and Thorne 2000). It performed useful and trailblazing service there until 1964. During CSIRAC’s time in Melbourne, the mathematics professor Thomas Cherry programmed it to perform music, developing a system and program such that anyone who understood standard musical notation could create a punched-paper data tape for CSIRAC to perform that music. Although the music performed by the CSIR Mk1 may seem crude and unremarkable compared to the most advanced musical developments of the time, and especially to what is possible now, it is probably the Ž rst music in the world to be performed on a computer, and the means of production lay at the leading edge of technological sophistication at that time. These Ž rst steps of using a computer in a musical sense occurred in isolation, but they are still interesting, because the leap of imagination in using the  exibility of a general-purpose computer to create music and the programming ingenuity required to achieve it are signiŽ cant. CSIRAC took some initial steps in that direction. An Overview of CSIRAC

Tune Retrieval in the Multimedia Library

Musical scores are traditionally retrieved by title, composer or subject classification. Just as multimedia computer systems increase the range of opportunities available for presenting musical information, so they also offer new ways of posing musically-oriented queries. This paper shows how scores can be retrieved from a database on the basis of a few notes sung or hummed into a microphone. The design of such a facility raises several interesting 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 and evaluate its performance.

The Development of the Notation Interchange File Format

This article presents the development history of the Notation Interchange File Format (NIFF), a glimpse into its design process, and a review of some of the format's main principles and features. A relatively new standard, NIFF is designed to allow efficient and comprehensive data interchange between programs that use common music notation (CMN). Such programs include editing and engraving programs used by music publishers, composers, and arrangers; sequencers, which allow music to be input via a MIDI keyboard and displayed using standard music notation; scanning programs that facilitate the input of large amounts of data; and specialized applications common to musicology and theoretical research.

Ballade, Patch and Sequence Editor for the Roland MT-32 Tone Generator and IBM PC

Ballade runs on the IBM PC, XT, AT or any compatible line of computers with 640 Kbytes of memory. Running on an AT/286 machine, Ballade has full capability, but with a PC or XT 8086 machine, there are limitations to real-time note display. This package is designed primarily for use with the Roland MT-32 Tone Generator, which is based on L/A synthesis or wavetable crossfading. Ballade will control all functions of patch, instrument definition, and sequence capabilities for the MT-32. For other tone generators it is only a ten-track sequence recorder. The required MIDI interface is a MPUIPC or MPU-401 with MIF-IPC or compatible. Three screens or modes of operations are offered to the user: (1) the Play Mode, with level meters and volume faders moving in real-time to indicate the level set for each track at any given playback moment; (2) the Song Mode, with three ways to input data-real-time recording, using the mouse to place notes on screen, or step-time recording; and (3) in Tone Mode MT-32, where you can define sound or patch parameters. Ballade supports recording, editing, and playback of the full MIDI event range. Ballade is a solid MT-32 tone editor and librarian. The ten-track sequencer matches the MT-32 track assignment, with track 1 for keyboard, tracks 2-9 for the MT-32 programmable voices, and track 10 for rhythm. The Play Mode screen has the look and functionality of a multitrack tape deck and mixer. The implementation of record, play, fast forward, and rewind buttons makes it easy to use. The sliders for each channel move with the mouse point, click and drag action and provides instantaneous response to the MT-32. Included in the Play Mode are panpot, mute, volume for each channel, and reverberation controls, which cannot be accessed by the front panel of the MT-32(!). In the Song Mode you can create songs by manually playing a keyboard, playing in the MIDI step mode, or entering notes on the standard music notation staff with the mouse. All notes and standard symbols can be displayed on staff lines. On the Song Mode page the Palette pull-down menu controls the display of slurs, dynamic marks, tempo changes, pitch bend, and pedal marks. Display of the Palette items is toggled ON/OFF with the mouse. Tracks 1-9 display on standard staff lines. Track 10 displays as a single rhythm line notation. The Song Mode has the usual cut, copy, paste, and erase edit options. One nice addition is a write function that lets you cut a segment of a melody and place it at any interval above or below another m lody; the cut and write can be from one track to another.

The Kyma/Platypus Computer Music Workstation

The Kyma/Platypus workstation is a graphic environment integrating sound synthesis, computeraided composition, and digital audio signal processing. All sound generation and processing is done in software, and each copy of the system can be modified and extended by the composer using it. Kyma is a language based on Sound objects; it is written in Smalltalk-80, running on an Apple Macintosh II. Real-time software synthesis is provided by the Platypus, a digital signal processor designed and built at the University of Illinois. The Kyma/Platypus framework is flexible enough to support such application areas as composition, systematic musicology, digital audio post-processing, sound track production, psychoacoustics, digital signal processing, the design of test signals, communications research, the sonic representation of scientific data-in short, any application requiring the interactive design of highquality digital audio signals. Three factors contribute to the flexibility of this system: (1) the choice of Sound objects rather than standard music notation as the representation for musical signals, (2) the choice of Smalltalk-80 rather than a compiled procedural language as the programming environment, and (3) the choice of a programmable digital signal processor (rather than a hardwired synthesizer) for the production of sound.

Using Connectionist Models to Explore Complex Musical Patterns

Music teachers often ask their classes to take musical dictation. The teacher commonly plays a short excerpt, perhaps a phrase from a Bach chorale, and then the students attempt to translate the various pitches and durations into the symbols of standard musical notation. Each resulting symbol can be judged true or false, right or wrong. Music teachers also ask their classes to analyze the musical syntax of phrases from Bach chorales. The resulting answers will likely have varying shades of truth or falsity. Students can have legitimate differences of opinion about how to interpret ambiguous musical events or events that have one meaning in prospect and a second in retrospect. And not only students differ in their assessments of complex musical patterns. Acknowledged experts in eighteenth-century music are known to disagree about how, in a given phrase, the chords, rhythms, contours, melodies, textures, timbres, and counterpoint all interact. Connectionist models offer elegant ways of dealing with multidimensional complexity of the type found in polyphonic, harmonically oriented music. One can, for example, define a musical event as an input pattern of activation that is then transformed by a network of interconnected processing units into an output pattern of activation representing an interpretation of the event. Models that allow for learning often require an explicit teacher prepared to state unambiguously whether, and to what degree, the output pattern is in error (Rumelhart, Hinton, and Williams 1986; Werbos 1988). For relatively low-level musical tasks, like the transcription problem described above, this requirement poses no serious obstacle; but for higher-level tasks, the notion of an explicit teacher becomes problematical. One wonders, for example, if anyone would be comfortable in claiming that one interpretation of a musical phrase is only 69 percent as true as another? An alternative class of connectionist models

The perception of metrical structure: Experimental evidence and a new model

This paper presents experimental results which are in conflict with existing theories of metrical interpretation [H. C. Longuet‐Higgins and C. S. Lee, Perception 11, 115–128 (1989); H. C. Longuet‐Higgins and C. S. Lee, Music Perception 1, 424–441 (1984); F. Lerdahl and R. Jackendoff, A Generative Theory of Tonal Music (MIT, Cambridge, MA, 1983); D. J. Povel and P. Essens, Music Perception 2, 411–440 (1985)] and a new model, based on the algorithm in Longuet‐Higgins and Lee (1982), which accounts for the main features of these results. Musically sophisticated subjects listened to sequences of identical tones in various rhythmic arrangements and transcribed them using standard musical notation. The results disconfirmed key predictions of all four theories, in that subjects displayed a general reluctance to choose interpretations containing an upbeat. However, the results did show that subjects were more prepared to avoid syncopated interpretations than those in which long notes occurred on weak beats. These tendencies are accounted for in the model by the assumption that the listener initially takes the first note of a sequence as the downbeat and is reluctant to abandon this (or later) hypotheses. But the occurrence of syncopations constitutes strong counterevidence and the occurrence of long notes on weak beats weak counterevidence, which can overcome the listener's conservatism. [Work supported by Nuffield Foundation, Royal Society and SERC.]

Dance to the Fiddle, March to the Fife: Instrumental Folk Tunes in Pennsylvania

The 651 tunes presented in this book represent an important segment of the musical heritage of North America and comprise what may well be the largest collection of Anglo-Celtic-American instrumental folk music. Tunes brought over by the first settlers from England, Germany, Scotland, and Ireland remained essentially unchanged despite many variations in the less accessible parts of Pennsylvania, thus preserving a musical tradition that stretches from colonial times to the recent past. The advent of electronic media and automotive transport brought an end to the isolation that sustained that tradition, and today its music is being diluted and homogenized through outside influences.Professor Bayard began collecting tunes in southwestern Pennsylvania in 1928 by listening to local fiddlers and fifers and transcribing what they played into standard musical notation by ear. His later collecting was aided by the addition of a tape recorder in 1948. The tunes he collected, along with those furnished by his associates in this project, were then sorted and organized into the present collection of 651 (some with as many as twenty-five versions) and annotated with the aid of leading authorities and published sources.The book starts with an introduction describing the general character of the tunes, the playing techniques of the fiddlers and fifers, and the circumstances in which the tunes were played. A note on the treatment of the music explains how tunes were used for different purposes (various dance and march steps, for example) through rhythmic modifications, and also details how fingering in traditional fiddling is related to key and pitch. The 651 tunes, arranged roughly by function (i.e., by the type of dance with which they are most often associated) are each labeled to show whether they were played on the fiddle, the fife, or both, and each tune version is identified by when, from whom, and by whom it was collected. An appendix connects the Pennsylvania tunes to other ones by giving transcriptions of forty-four related tunes from Missouri, Indiana, Massachusetts, Prince Edward Island, and County Cork, Ireland. Biographical sketches of the performers, a list of sources and an index of tune titles conclude the book.The authentic tunes of the Bayard collection will be invaluable both to the folklorists and to composers, performers, or students of music who seek understanding of the American idiom. Notes following each of the transcriptions link the tunes to others of Anglo-Celtic provenance, as well as to continental European airs.

The New Harp of Columbia

In the small towns and rural areas of early America, church-sponsored singing schools proliferated as a way of both improving congregational singing and drawing communities together. Congregants attending these schools were taught a form of musical notation in which the notes were assigned different shapes to indicate variations in pitch a method that worked well with singers having little understanding of standard musical notation. These schools eventually became major social events that drew hundreds of attendees, and today countless enthusiasts carry on the shape-note The New Harp of Columbia, originally published in Knoxville in 1867, was a shape-note tunebook used in East Tennessee singing schools. It was based on an even earlier publication, The Harp of Columbia (1848). In 1978, the University of Tennessee Press published a facsimile edition of The New Harp with an introduction by Dorothy D. Horn, Ron Petersen, and Candra Phillips that detailed the history of shape-note singing as well as the story of the tunebook itself and its original compilers, W. H. Swan and M. L. Swan. That edition went out of print in 1999. Now, for this restored edition of the tunebook, the Press has reprinted not only the full text of its 1978 facsimile edition but has included additional tunes that were part of the original 1848 Harp of Columbia. A few verses to some songs favored by contemporary singers have also been added, and a new foreword by Larry Olszewski and Bruce Wheeler brings the story of the tunebook and its users up to date. Included in the book are old psalm and hymn tunes, anthems, fuguing pieces, and folk hymns a total of more than two hundred pieces that represent a fascinating slice of Americana. As a reviewer for the Journal of Church Music noted of the 1978 facsimile: [The book is] a worthwhile addition to any church musician s library, especially those interested in the development of American sacred music over the past two centuries. This publication marks a significant new step in preserving an important musical tradition.

The perception of music.

A common but none the less remarkable human faculty is the ability to recognize and reproduce familiar pieces of music. No two performances of a given piece will ever be acoustically identical, but a listener can perceive, in both, the same rhythmic and tonal relationships, and can judge whether a particular note or phrase was played out of time or out of tune. The problem considered in this lecture is that of describing the conceptual structures by which we represent Western classical music and the processes by which these structures are created. Some new hypotheses about the perception of rhythm and tonality have been cast in the form of a computer program which will transcribe a live keyboard performance of a classical melody into the equivalent of standard musical notation.

The Perception of Music

AbstractThe author recently described elsewhere a computer program which would transcribe classical melodies played on an organ console into the equivalent of standard musical notation. This program was the fruit of a prolonged effort to understand how Western musicians succeed in making sense of music, in discerning rhythm and the tonal relation hip between notes. The interest of the problem arise from the fact that no two performance of the same piece of music are ever identical so that the listener has to discriminate between those variation of timing and pitch which are structurally significant and those which are merely expressive. In order to under land the ability of some musician to reproduce the cores of music that they hear it is necessary to develop a formally precise theory of rhythm and tonality – a theory which is couched in terms reminiscent of Chomskyan linguistics. But it is also necessary to consider how the listener builds in his mind a representation of the rhythm and tonality of perfo...

Perception of melodies

A computer program has been written which will transcribe a live performance of a classical melody into the equivalent of standard musical notation. It is intended to embody, in computational form, a psychological theory of how Western musicians perceive the rhythmic and tonal relationships between the notes of such melodies.