Neumatic

Neumatic is a suite of online software based around a musical notation format called Neumat. Neumat has been designed as a powerful, flexible and easy-to-learn file format and input method for Early Music notated in neumes.

Development of the notation and the software began only in November 2011, and everything is still very much at the prototype stage. This page will tell you all about current progress and plans for the future.

• Application suite
• Example score
• Neumat notation basics
• Neume examples

Application suite

There are three main components to the Neumatic software.

Neumatic NeumePad

A visual editor for inputting Neumat notation, designed to make the transcription of musical manuscripts written in neumes as fast and simple as possible.

Neumatic Engraver

A generator of high-quality scores from Neumat notation in a variety of file formats (expected to include PDF, SVG and PostScript). Note that NeumePad, however pretty it may look on screen, is purely an input tool: it is not, and never will be, intended for final output. Neumatic Engraver does much more, the most important thing being to combine the musical notation with the lyrics as seen in the sample below. It also handles line breaks intelligently, justifies staves, adds titles, and does all of the other good stuff that is important for presentation but doesn't matter at all when you’re transcribing the manuscript to begin with.

The Engraver is not yet ready to try out, sorry, but I hope to make it available here soon.

Neumatic Converter

A set of tools for conversion of Neumat scores to modern notation, including the assignment of mensural values to neumes. I haven’t even begun development of this yet, but watch this space!

Read on below for more details of the Neumat notation format. If you'd like to know more about current progress with the development of the software, feel free to drop me a line at info@cantigasdesantamaria.com.

Example score

Here’s a score rendered by the prototype Neumatic Engraver, combining the music for Cantiga de Santa Maria 383 taken directly from the Escorial j.b.2 codex, with the edited lyric text from my Cantigas de Santa Maria for Singers website:

Neumat notation basics

In Neumat notation, the music above looks like this:

Although this might look like gibberish at first sight, it’s actually a very simple format which quickly becomes familiar. Here are a few of the most important features:

• It’s just plain text, making it easy to read and write for both programs and people alike.
• The notation is made up of a simple sequence of elements separated by whitespace. The amount of space doesn’t matter as long as there’s at least one space, tab or newline character, so you’re free to lay it out as readably or as messily as you please.
• Most elements begin with a positive or negative number that represents the vertical position of the element on the stave. Zero indicates the middle line or, if there are an even number of lines, the line just below the middle. This brings us right away to an important point about the Neumat format, which is that it is first and foremost a representation of what the stave looks like, not how the music sounds, and so the actual pitch of notes is only coded indirectly. (If you’re familiar with neumatic notation, you’ll know already that the clefs were never fixed in position but free to roam up and down the stave within a single score, so a given line doesn’t even have a fixed relative pitch, let alone an absolute one.)
• If the number is missing it’s understood to be zero, but in practice this is only done for the few element types which have an invariable vertical position anyway, such as the boxes and lines which delineate musical sections and phrases, or the sawtooth which encodes a line break in the manuscript.
• Apart from this optional position, element codes are made up of sequences of letters or a few special characters.
• A subset of the element codes are ‘simple’ with no sub-structure. This includes * > | " and ' which indicate divisions and breaks:

  *	start of section (refrain, stanza)		Rendered by Neumatic Engraver to contain the first letter in the section, like a decorated capital.
  >	manuscript stave break
  |	metrical phrase break
  "	double small separator
  '	single small separator

Manuscript stave breaks are generally of no musical consequence, but they are usually followed by clefs (see below) which obviously are. The reading of the phrase breaks and other separators as pauses or rests is definitely a question of interpretation, and is not really relevant to the visually-based Neumat specification.
• The ‘simple’ elements also include the clefs c and f, and the B-flat accidental @:

  c	C clef (do)
  f	F clef (fa)
  @	B-flat accidental

  Note that the correct vertical position for @ is fixed by the position of the c or f which precedes it, so it shouldn’t be written separately in the Neumat code. The c and f clefs themselves need a position, however, as they can move around freely (in most cases the position should correspond to a line on the stave, not a space).
• The last ‘simple’ element is the ? for a query (also known as the ‘WTF element’). This very useful code can be used during transcription to temporarily mark an element in the manuscript which you're not sure you've read correctly and will check later:

  ?

  query

  In a future version of Neumat I’m intending to add further ‘editorial’ codes, in particular so that obvious scribal mistakes in the manuscript can be marked up with suggested corrections, in such a way that the original symbol is also preserved along with its replacement.
• All the other element codes, such as o, on, owo, ra, eyeye, bowowoyo, baron or ohodra in the example above, are ‘compound’ and represent neumes. The individual component letters or ‘codons’ here have specific meanings for bodies, stems and pitch steps, and can be used in a multitude of different combinations to build up the huge variety of neume shapes found in the manuscripts, from just a few simple parts:

  o	square body		Every neume must have at least one body part. As a special case, two a codes in a row combine into a single body part that descends two steps, rather than being two separate bodies each with a single step.
  a	descending body
  e	diamond body
  b	top-left stem		Each body part can have zero to four stems, but more than two is unusual in practice.
  r	bottom-left stem
  n	bottom-right stem		Right-side stems are automatically turned 45° when added to a diamond body.
  d	top-right stem
  x	strike through preceding stem		For corrections appearing in the original manuscript.
  w	step up		These are relative steps up or down the stave between subsequent parts of the same neume, as distinct from the number before the neume (see above) which fixes the position of the first part.
  y	step down
  s	stacking step		A step up and back, where one body sits directly on top of another, e.g. in the podatus neume (see oso in examples)
  h	(optional code) no step		Does nothing (see note below.)

  You might notice that although the resulting neume code strings (owo, ra, eyeye, bowowoyo, ohodra etc.) are strange to look at, they are all more or less pronounceable. This has been a deliberate design decision in creating Neumat, and it aids learning and memorability enormously. After just a few minutes playing with neume codes, you’ll find that you can create any new shape you want with just a moment’s thought, and those that are most commonly used (on, ra, ron, owo, etc.) quickly become second nature. (NeumePad also provides convenient keyboard shortcuts for even faster entry of basic neumes, such as q for the virga (on), but these don’t form part of the Neumat specification.) Note, by the way, that the code h doesn’t actually do anything, and is provided solely to assist pronunciation where two body parts would otherwise be juxtaposed, e.g. in ohodra.
• Since a single neume may have multiple bodies, there are rules concerning the order in which the parts must be entered, in order to avoid ambiguity. The main constraint is that a left stem must be written before the body it belongs to, and a right-stem after it, but as this is rather intuitive there’s generally no problem in learning the rule, and once the common shapes are typed a few times, it becomes a complete non-issue. It’s also obvious when things are in the wrong order, as the neume code will most likely be unpronounceable!

Neume examples

In the image below you can see a selection of more than 70 different neume shapes sampled from the first 70 or so Cantigas de Santa Maria from the Escorial j.b.2 codex, all with their Neumat codes. Hopefully this will demonstrate at a glance the simple power of Neumat. You can also see that all the combinations are just about pronounceable, or at least they are if you’ll admit w as a vowel once in a while, as in onwba (no problem if you’re Welsh).

As a challenge, once you've digested the summary above, have a go at covering up the codes below and see if you can work out what they should be. It's easier that you might think!

Admittedly, at this early stage of development of Neumatic I’ve concentrated very much on the mensural neumes in the Cantiga manuscripts with which I am most familiar, and there are certainly whole load of other manuscripts in neume notation that the system cannot yet reproduce fully. Rest assured I’ll be working on it though!