Beyond MIDI: The Handbook of Musical Codes

---

The examples of musical code shown in this book are based on a common set of short notated examples which collectively illustrate some of the most common problems in musical encoding generally and which also tend to highlight the differences between sound-oriented approaches and notation-oriented ones. The number of examples set was left to the discretion of each contributor, and would inevitably be influenced by the purposes of the code, the current state of development, and the time available to the contributor.

Each example is presented and discussed below. The functional titles used are also retained in the chapter descriptions. Many contributors submitted output as well as code, but since the aesthetic merits of different systems are not our topic here, we have standardized the presentation: Examples 1.1, 1.2, 1.3, 1.4, 1.5B, and 1.6 were set by Nicholas P. Carter using SCORE; Werner Icking's setting of Example 1.5A is used in Chapter 17, where numerous special characters have also been produced using MusiXTeX; in this chapter Example 1.5A is scanned. 4.1 The Mozart Trio

This example, which is taken from the second trio section of Mozart's Clarinet Quintet, is musically very simple. Respondents were asked to set only the first twelve bars (up to the repeat sign).(17) Respondents for monophonic systems were asked only to encode the Clarinet in A.

The entry was intended to be simple. Each instrument occupies a different staff and plays single notes. The meter is regular, although there is a tuplet in Bar 8 of the clarinet part. The articulation marks include slurs and staccatos.

The most telling feature of this example is the transposing part of the Clarinet in A, which is notated in the key of C Major, three half steps higher than its sounding pitch of A Major. Although transposition is a simple matter on the computer, the vertical shifts in layout often require that notation users re-edit the graphic result. Although in Example 1.1 the space between staves is generous, it often happens that low notes on one staff will overwrite high ones on the staff below or dynamics signs between staves or, in vocal music, text underlay beneath the part.

The implementation of sound applications is not subject to such considerations. In sound applications, however, one may wish to consider the consequences of the repeat signs, for what is only 12 bars in print could, if these are acknowledged, become 24 in sound applications. If the example were of Baroque music, the software developer might wish to leave the option to repeat the passage up to the user by providing a switch of some kind, but repeats in Mozart minuets (of which this example is a part) are not regarded as optional. Example 1.1 Second trio from the Mozart Clarinet Quintet, K. 581 ("Mozart trio"). 4.2 The Mozart Piano Sonata

In the identification of staves and systems, piano parts are atypical. One "instrument" typically occupies two staves, and within the context of a work for multiple performers these two staves may function as one sub-system. This example exhibits some other complications: (a) mixed durations within single chords in the right hand, (b) grace notes preceding chords (right hand), and (c) arpeggiated grace notes (left hand).

Reconciling these arpeggiations with those of the right-hand chords in real time is a task that none of our contributors sought to explain. Example 1.2 Excerpt from the Mozart piano sonata in A Major, K. 331. 1.4.3 The Saltarello

This anonymous dance piece from the early Renaissance is very simple musically. The complexity it represents is one of processing sequence, on account of its multiple endings. In this case a representation for printing only is significantly simpler than one designed for playback or structural analysis. Example 1.3 Anonymous saltarello. 1.4.4 The Telemann Aria

This aria is scored for voice (Staff 1), an instrumental ensemble consisting of oboe, violin, and viola (Staff 2), and a basso continuo complement of cello or double bass and keyboard (Staff 3). The right-hand part for keyboard is not realized here.

The vocal part contains text underlay that includes the need for diacriticals (here the ö and ü) and special characters (German ß). The width requirements of the text syllables are extremely erratic, complicating text underlay. Example 1.4 The Telemann aria "Liebe! Liebe! Was ist schöner als die Liebe?".

The information on Staff 2 is very complex. The oboe and violin generally share the same part (shown in full-size notes), although they split in accompanying the words "süsser" and "schöner." The viola part (shown with stems down) contains a mixture of single notes and double stops. In Bar 10, there are mixed durations (quarter and eighth notes) in the double stop on the second beat. The cue-size notes are editorial ones suggesting material that is missing in the original source.

These features make sound interpretation from a representation devised for printing particularly troublesome. The oboe and violin require separate sound tracks because of their very different timbres. In these separate tracks the notes will not be identical. The execution of the trill in Bar 10 also requires the output of sound in multiple tones in real time where a print file contains only a single symbol.

The basso continuo part (Staff 3) is more straightforward, but it also requires some differentiation of temporally different parts on a common staff.

If a sound representation is taken as primary, the task of achieving this visual grammar without significant amounts of hand-editing is daunting. 4.5 Unmeasured Chant

Two examples are given. "Alma Redemptoris Mater" (Style A) is shown in neumes on a four-line staff as it appears in the Liber Usualis. Example 1.5A Chant: "Alma Redemptoris Mater" written in neumes.

"Quem queritis" (Style B) is shown in modern notation that offers phrase markings but does not impose any rhythmic identity on the notes. Example 1.5B Chant: "Quem queritis" written in modern notation.

Electronic-sound representations of chant are almost meaningless, since exact durational information was not precisely conveyed and the interpretation of relational durations remains a subject of debate. 4.6 The Binchois Magnificat

This three-voice work by the Renaissance composer Gilles Binchois illustrates some of the complexities of distinguishing original from editorial material in representations of modern editions of early music. On Staff 1 a single voice (the Cantus) is notated. On Staff 2, System 1 two voices (Cantus 2 and Tenor) play a common instrumental part (stem directions are determined by proximity of the notehead to the middle line of the staff). On Staff 2, System 2 these instruments have separate parts (indicated by upward stems for Cantus 2 and downward ones for the Tenor). Example 1.6 Binchois Magnificat.

The main complexities of this example are in the editorial markup. Editorial elements include (a) the incipit of Cantus 1 in "colored" but unstemmed notes, (b) horizontal brackets showing where ligatures(18) were used, and (c) editorial accidentals (e.g., in Bar 6).(19) 4.7 Free Choices

Contributors were invited to submit additional examples to explain their treatment of contemporary music or other repertories outside the realm of "common music notation." There were very few contributions of this nature and no way of comparing them, as the range of printing aberrations is great. The concurrent use of multiple meters, unusual noteheads, beams unconnected to stems (and therefore to notes), wedge-shaped beams, and circular staves are some of the features that have marked contributions shown in recent issues of Computing in Musicology, particularly Vol. 9 (1993-94). Numerous special signs are also required in popular music repertories. Footnotes

17. Some encodings also necessarily capture the layout. The example was distributed in two different layouts--one with a system break after Bar 7 and the other with a system break after Bar 10.

18. Contiguous or composite symbols representing two or more notes in the original source; such visual arrangements conveyed information about performance such as phrasing and, to a degree, duration.

19. In the Renaissance practice of musica ficta, some accidentals were implied by the hexachordal modes used in composition but they were not made explicit in notation. Modern editions tend to make them so. Not all editorial accidentals are accepted by performers, who therefore wish to know which accidentals are added and which are original.