There will be, at some point, a post on the method used to generate some of the pitches for this piece. For this I used an (shock, horror) algorithm.

The instrumentation is:
string quartet
brass quintet
piano, laptop, soprano, electric guitar, and drum kit

The Soul Weeps – Movement 1 by joshkopecek

Another video, this time of Psappha performing The Warrior Fallen.

Promotional video for Ninnananna.

NINNANANNA from made in apu on Vimeo.

I think it is necessary to look beyond this, and to think about intervals, major and minor harmony, scales, dodecaphonicism, in a new light. One where the relationships remain constant for harmony across all styles. It’s useless to describe Boulez’ music in terms of major and minor, as with Ockeghem but for completely different reasons.

So, he comes to the harmonic series. This is the one constant in music, present in all pitched instruments (in fact the basis of how we recognise pitch), and very possibly the key to how our ear is able to distinguish two separate pitches in a chord, even when they are almost in unison. For my Masters I did some research into dissonance, and made some inroads into how pitches interact, and for example why an octave is considered consonant by most people and why a minor second is dissonant. Basically, although inconclusive, octaves share most harmonics with each other, apart from one per octave. Minor seconds, on the other hand, share very few, at least until many octaves higher in the harmonic series, when the harmonics become less than a semitone. I set about the work of providing a theory of this, which is still in progress.

More recently I came across (whilst reading Michel Chion’s translation/transmogrification of Schaeffer’s Traité) Edmond Costère’s Lois et Styles des Harmonies Musicales. Costère’s book describes his reexamination of harmony from the perspective of basic rules of acoustics, primarily the harmonic series, and provides charts of analysis of scales from this perspective, an analysis of major harmony, it is disarmingly simple, and was just as easily dismissed at the time of publication. But it is quite obviously before its time, before Spectralism appeared and when people were still grappling with the Highest Art preoccupations of whether harmony needed reinventing or not. Costère is not suggesting we restyle everything, but in fact to reexamine the way we look at harmony, that there is a bigger picture and a simpler but more comprehensive method of analysis of functional and non-functional harmony, that spans the gaping abyss between tonal harmony and serial music, between spectralism and new complexity.

What is less convincing for me is his explanation of the roots of diatonic minor harmony, which consists in inverting the harmonic series in order to find the minor third. Instead for me at this early stage I see minor harmony as a séquelle from the modes; diatonic minor is quite easily accounted for as the aeolian or dorian medieval modes, shoehorned into functional harmony (sharpening of the 6th/7th degrees). Although, this doesn’t fit it in quite so neatly into Costère’s hypotheses.

Currently, this is an unfinished story. I have not finished Costère’s tome, nor the other equally outrageously titled Mort ou Transfigurations de l’Harmonie, which I have on loan from the British Library. I intend to assimilate, modify and bring his methodologies for analysis and appraisal of harmony into the 21st Century. It will become a tool for analysis applicable to most systems of harmony (diatonic, dodecaphonic, quarter-tone), and possibly may influence the way I compose, harmonically speaking. This has, of course, nothing to do with rhythm, timbre, structure or form, thus leaving one free to compose the same music but aware of another perspective on the harmony that they are using. Suffice to say that for me this is a breakthrough, rather late in the day, but it has answered some questions that I have long posed (those of a quantitative scale of dissonance and alternative hierarchy of tones) and equally triggered others (was Costère just a blithering fool with a prejudice against established rules of harmony?).

The concert starts at 7pm New York time, which is midnight (00:00-02:00) Saturday 19th December if my calculations are correct.

The concert with be streamed live here, so tune in if you can. Click here to add this to your calendar.

There are new pages for my catologue of works – Music; Ninnananna – with pictures from performances and information; an About page with my contact details; this blog with updates on my research; and the Events page upcoming performances.

When I first started working with Ruby and Cocoa, the RubyCocoa project was pretty elementary, but many leaps and bounds have been made. Thanks go out to bleything and and the RubyCocoa team without whose wonderful work this would not be possible.

Making an interface from scratch with the Arduino at its core

Equipment needed:

• arduino (diecimila or equivalent)
• wire + wire cutters
• breadboard
• selection of resistors (10k preferable)
• switches (any sort, I used momentary)
• computer – mac, linux or pc (mac preferable)
• box of some sort – optional
• soldering iron – only necessary for finishing the project

# Wire a switch – one side to ground and the other to digital input 6. Use the breadboard.
# Put a resistor between the side of the switch connected to ground and +5v.
This is called a pullup resistor and makes the current default to ‘on’ or +5v when the switch is closed. When the switch is open, the voltage remains at 0v.

If you want additional switches, follow the two steps above making sure you use a new resistor each time.

Difficult bit: find a nice box and glue the whole lot into it and plug in a usb cable.

Getting your computer prepared to talk to the Arduino

If you are comfortable with using the terminal, and want a one-step install for the various packages required for this tutorial, try:

MacPorts

To install: go to the MacPorts website and install using the installer there.
Check everything is working by firing up a terminal and typing

port version

which should return

Version: 1.710

Preparing Ruby

All the below are necessary before continuing

Installing (upgrading) Ruby
Ruby is already included in MacOSX, but for compatibility reasons, it’s best to upgrade. If you’re happy with your installation, just continue as is until you come up against any errors. Please note the Ruby included with 10.4 is very old, and possibly broken.
Fire up a terminal (if it isn’t already open from above) and type

sudo port upgrade ruby

Rubygems — required to use the below libraries
Fire up a terminal (if it isn’t already open from above) and type

sudo port install rb-rubygems

MIDIator — Ruby MIDI library
MIDIator allows us to use Ruby to easily create MIDI messages in a simple and comprehensible fashion
Fire up a terminal (if it isn’t already open from above) and type

sudo gem install midiator

Serialport — Ruby Serial Port library
This is installed by default as part of Ruby, but best to upgrade:

sudo gem upgrade ruby-serialport

Easy! Just sit back and watch incomprehensible lines of text scroll by.

Arduino code

The Arduino itself requires code to run. This tells the Arduino what to do with incoming and outgoing data. The language is called Wiring.

Make sure you have the Arduino programming environment installed (from http://arduino.cc)

Use the following code:
To explain, the first block initialises the pins we will use.
The setup() method initialises the pins and serial connection and tells us that they are sending data in to the Arduino.

The loop() method reads the pins every cycle, stores the data in variables and decides what to do with it. In this case, it sends a different value to the computer if the pin is up or down. The swi variable checks if it’s already been sent, and prevents doubling the ‘note on’ value. i.e. it can’t send a note off before it’s sent a note on and vice versa.

 /*

 sends serial data over usb to be converted by ruby to MIDI..

 */

int inPin1 = 6;   // choose the input pin (for a pushbutton)
 int inPin2 = 7;
 int val1 = 0;     // variable for reading the pin status
 int val2 = 0;
 boolean swi1 = false;
 boolean swi2 = false;

void setup() {
 Serial.begin(9600);
 pinMode(inPin1, INPUT);    // declare pushbutton as input
 }

void loop(){
 val1 = digitalRead(inPin1);  // read input val1ue
 val2 = digitalRead(inPin2);

if (val1 == HIGH && swi1 == false) {         // check if the input is HIGH (button released)
 swi1 = true;
 Serial.println(89, DEC);
 delay(100);
 }

if (val1 == LOW && swi1 == true){            // check if the input is low and it's just been high
 swi1 = false;
 Serial.println(90, DEC);
 delay(100);
 }

if (val2 == HIGH && swi2 == false) {         // check if the input is HIGH (button released)
 swi2 = true;
 Serial.println(91, DEC);
 delay(100);
 }

if (val2 == LOW && swi2 == true){            // check if the input is low and it's just been high
 swi2 = false;
 Serial.println(92, DEC);
 delay(100);
 }

}

Building a simple communication script in Ruby

Download the RubyMIDI.zip file and look in the lib folder at usbchecker.rb. This checks to see if the Arduino is present. In short, OSX creates a list of all the USB devices on your computer at /dev/. This script checks for one in the correct syntax for an Arduino, and passes it back to our main program.

Look in the main folder at BasicMIDI.rb. This contains all the information to communicate with the Arduino and change it to MIDI data.  The require functions at the top load in the libraries we installed earlier – rubygems, midiator, serialport and the file we’ve just looked at, usbchecker.rb. The first paragraph runs the usbchecker.rb and passes the identity of the Arduino back and initialises it (turns it on and starts communicating with it).

The next paragraph creates a MIDI device with Midiator and tells it to send all MIDI data to CoreMIDI on OSX, also known as the IAC Driver (Load up Audio MIDI setup and have a look).

The whole of the next section is a bit boring, and merely consists in grabbing data from the serial port and converting it into a meaningful format rather than the incomprehensible rubbish that comes across normally. sp.getc.chr grabs a character (getc) from the serial port (sp), and converts it into a real character (chr). The next loop puts it into a string.

Then, if it has something, it decides which one it’s received from the Arduino (see above that we’re sending different numbers according to what’s happening) and sends either a midi note on (midi.note_on(89, 1, 100)) or note off (midi.note_off(89, 1, 0)).

The other parts are failsafes for if it receives a line-end or newline character – the Arduino sends these after each transmission as part of the Serial.println() method.

Magic, that’s all the code explained

We can run this script by opening a terminal, typing ruby and dragging and dropping the BasicMIDI.rb file into the window, then hitting return. It will fail if there isn’t an Arduino present.

Integrating your Ruby script into a fully fledged native MacOSX application

First, go and download Xcode from http://developer.apple.com and install it. This will take a while.

You’ll notice there’s another folder called RubyCocoaUSBApp which contains lots of files and folders and a file ending .xcodeproj which has our Xcode project in it. Open this, and Xcode will open the project. You’ll notice the lib folder is still there, and that AppController.rb has the same code as BasicMIDI.rb above. It’s not necessary to know the structure of the rest of the folders, just that there’s lots of necessary stuff in there that it’s unwise to touch. Just modify the AppController.rb with your code and link it into the MainMenu.nib using Interface Builder – I won’t explain here because it’ll take ages, unless people need me to? Easy peasy.

Next step: Build and Run (Command-R).
This should launch our fully fledged application, which runs when we click on the button. It will display ‘Ready’ when connected and sending MIDI.

Simply connect any MIDI client to the built-in IAC Driver on OSX and away we go.

Oh, and before I forget, here’s the fully fledged application in case you’re having problems compiling using Xcode.

Firstly I think it would be a mistake to think that this paper is about time in music. The word time, when brought up in relation to music practice, usually recalls such words as ‘tempo’, ‘rhythm’, maybe even ‘groove’. However, although Grisey starts by talking about rhythm, he’s referring to the experience of music. Jonathan D. Kramer also talks about the same thing in more detail in his book The Time of Music: new meanings, new temporalities, new listening strategies (1988). Both of these are an attempt to justify or somehow examine the difference between time as experienced in normal life (clock time) and time as experienced in music (musical time). I must say I’m not a fan of his confusing categorisation of the different types of time (the skeleton of time, the flesh of time, and the skin of time), as they are hardly as elucidating as the rest of the text.

To delve straight into the text, I’d like to reinforce an element in my thinking that Grisey raises, which I call repetition, and he calls periodicity. We are examining the same concept from different angles, mine from the ‘material’ point of view and he from the ‘rhythm’ point of view, although with a certain leap of faith we could regard them as interchangeable (rhythm being the repetition of material). I quote Grisey quoting Abraham Moles “the notion of rhythm is linked to that of expectation”, and this in turn leads us to realise that a framework has to be established in order to perceive rhythm. Repetition is that framework – if we repeat a bar twice, we expect it a third time. If we have three bars of material and we repeat the first two, we will expect the third. If we play quavers in 4/4 for 10 bars and switch to crotchets, we gain the listener’s attention when we change. Grisey returns to the concept of periodicity at the end of the paper:

How many times have we sat through a piece that had beautiful material in, but an hour after the concert was completely lost from memory? Sometimes, as a composer, it is forgotten that the listener may only hear the piece once. We’d be very lucky if they sought out a recording of the piece (if one was even available), and even then there is a limited amount of listens a particular person is going to dedicate to it. What I am getting at is that there is a limited (set) amount of time available to develop and demonstrate your material. We are also assuming that our dutiful listener is even paying attention the whole time through the performance, that it’s not boring them or that they’ve even had to sit through five other difficult, new contemporary music pieces.

Coming back to periodicity, Grisey outlines his composition Périodes, in which he examines fuzzy periodicity. This is the idea of a constantly fluctuating pulse, determined by (statistically-speaking) a sort of normal distribution around a fixed pulse. I challenge this notion as a compositional parameter. A performer can never absolutely determine the time between actions, only estimate. So even the strictest tempo from a performer constantly fluctuates around an ideal tempo. So what Grisey is suggesting is already part of performance. True, what he is really saying is that he wishes to investigate this phenomenom, but the truth is that the contrast between the intentional fuzzy periodicity and real performer leeway is much more subtle than he imagines.

In the acoustic world nothing is exact. Everything is an approximation—even frequencies, partials, rhythms, tunings, scales etc. In the electronic world, introducing this human element—the almost random/approximate/fuzzy—becomes a lot of hard work and is not intuitive. Often we have to try to approximate live performance by reproducing certain elements of approximate rhythm, timbre etc., and even more frequently is this element completely ignored. The perfection of electronic sound is immediately identifiable if not directly describable. We can listen to a piece of music and identify the parts that have been sequenced (read: never been near an instrument) and those which are recorded from real life. Likewise the ear will immediately notice an exact loop and be less fatigued by an indirect repetition. That means more mileage out of the same material: not boredom but a chance to convey our point more clearly.

To recapitulate or not to recapitulate. Recapitulation has been a mainstay of composition for centuries, for the reason that the listener immediately identifies with and gets a (small) sense of achievement from having identified material that was presented earlier. This immediately links the composer, performer and listener in a chain of understanding, and probably more importantly, communication—even on the most basic level. For me it seems that without this basic premise the music has to be stylistically immediately identifiable. That is, the audience is going to the performance knowing the way in which you work and the boundaries of your compositional language (your ‘sound’). They are prepared to be surprised, but they have a grounding upon which to base aural judgements.

There are some particular moments of insight, which I’ll work through in order. Firstly she cites Kandinsky: “Form is the external manifestation of inner meaning” (Vassily Kandinsky: The Spiritual in Art, 1969). Aside from the fact that musicians, musicologists and above all composers tend to take reference from out-of-date thinking in other art-forms, I think this is particularly apt, even now, as a guidance for how to approach form in music.

Too often is structure confused with form. Look at it this way; the way a piece is structured in terms of harmony, orchestration and material is completely separate from the way these elements are formed. To coin an analogy, form is the father of structure; at least in the classical sense. We have a set of boxes (possibly within larger boxes)—the form—which we fill with material—our structures. These boxes might even be predisposed towards a particular type of material. These constraints often paradoxically provided a composer with inspiration, as a kind of competition to perfect their art, in the manner of an acheiveable target.

What Kandinsky is suggesting is that there is, in fact, no box at all. Which is, in itself, obvious—there cannot be form without content. That the content shapes the form is the interesting part, because this suggests that in music the structures themselves do not hold the meaning, but instead the method in which they are arranged carries the semantic weight. Saariaho suggests that she has never referred to pre-established formal structures, and in fact it would be crazy to expect that a composer necessarily adopted forms developed directly from analysis.

Kaariaho goes on to say “Amongst familiar organizational models concerning pitch the tonal system is, in my own experience, the most effective means of suing harmony to construct and control dynamic musical forms.” and “I think, however, that using tonal functions in such a way is definitely a thing of the past. This is why the tonal systems would seem to be only a potential model for the creation of tensions through the use of pitches” and finally “…I am absolutely convinced that [tonal harmony] is an out-dated approach to the problems posed by the organization of pitch structures”. She goes to say that tonal harmony is the only way of effectively creating dynamic structures, but that perhaps it’s not necessary to do that anymore. Why do I raise this? Because I think this is an issue which has still not been satisfactorily solved.

Tonal structures are a proven, effective way of creating tension and meaningful structure in music. Serialism has left many unconvinced, in the same manner as Marxism: fundamentally sound but disastrous in practice. Musically speaking it let us down on several points: the matter of ‘octave equivalency’—notes do not sound the same in each octave; interval equivalency—a major third is not a minor sixth; set-equivalent chords are not identical sonically. Spectralism solved some of these problems by reintroducing the natural hierarchy of tones without reverting to tonal structures. Many of the issues raised with spectralism have been circumvented/solved by the advent of computer processing and analysis technologies (a contentious issue, I know), through the endless possibilities for manipulation of sound.

Saariaho sways towards investigating a hierarchy of timbre to replace tonal structures, but in my field I am swayed towards a complex appreciation of the interplay of tones. This is driven by my interest in the contradiction between the harmonic series and the system of equal temperament: two systems in use at the same time, contending for the same position. The physical irrefutability of the harmonic series leads us to appreciate the tonic–dominant–sub-dominant relationships. The dominant resolves strongly to the tonic, being part of the harmonic series of the tonic, and the modulation to the sub-dominant provides a leverage (alternate dimension?), because the tonic is a part of the harmonic series of the sub-dominant (and not vice-versa). The build-up of tension in the dominant is due to the third of the dominant chord wanting to directly resolve (up) to the tonic again, whilst this is not the case with the sub-dominant. Whilst these are very simplistic observations, it highlights one of the basic issues regarding the organisation of a hierarchy of tones and chords into a harmony, and how little we can actually categorise without resorting to a subjective use of our ear.

Saariaho also goes on to say “It is by modifying the speed of harmony’s development that harmonic tensions are created, by making harmony into a strictly controlled parameter. To achieve this, I worked on the degree of differentiation of successive chords an don the linear duration of each chord”. This is the eternal compositional tool of stasis vs. movement. Another is the level of interest generated by material. One can have very mediocre material juxtaposed by inventive (catchy, if you will) material, in order to heighten its catchiness. This is the method unconsciously suggested by thematic development, that one suggests imperfect combinations of what one has heard before, ultimately recapitulating on the initial material, satisfying the listeners’ desires. An even level of release of good material would result in a saturated work. The listeners’ ears need to be teased, like in a horror movie when the tension is sufficiently developed.

I was investigating another parallel area of interest, that of Bill Viola’s work in visual art. To describe his work here in such a short manner does not do its depth of thought justice, but he (currently) works with extremely slow video, in the manner of a very subtly animated painting. The casual viewer would assume a light-box photograph or extremely high-quality video still, but in fact it is an extremely slow video. The concept of going beyond the medium of video to question its very premise of presenting motion is one to be greatly praised. I am considering how this could be possible in music, to present a moment of stasis in a particular musical texture.

The normal way we elongate events is to literally play them more slowly, and the electronic equivalent is time stretching using a phase vocoder, preserving the pitch. Under normal circumstances the usefulness of this (without disturbing the formant arrangement of the original) is limited to twice and half the length. The factors that have to be taken into account are the numbers of bands it is split into (powers of 2) window overlap (again powers of 2) and window shape. The most low level process I have found is the Csound and Matlab implementations, which might provide a little more room to maneuver. Also band-splitting the sound file into (at least) high and low might preserve individual formants better with extreme processing.

Saariaho examined this principle in Vers le Blanc (1982), in which a single extremely slow glissando from one chord to another lasts 15′, subverting the perception of harmonic change. Leif Inge examined the same principle with 9 Beet Stretch, Beethoven’s 9th Symphony stretched to 24 hours long. There’s also John Cage’s ORGAN2/ASLSP, currently being performed in St. Burchardi in Halberstadt, in the eighth year of its 639 year performance. So the idea of performing music for a very long period isn’t new, but Cage’s doesn’t examine slow, continuous change, rather specific moments in time that have sudden changes. Also it’s not feasible to truly experience the work, because the time spent in performance for one human wouldn’t be enough to experience the enormity of the work. We can only imagine the momentous moment of change.

In Saariaho’s work, the computer is treated as an assistant to composition, not as a replacement for any part of the composition process. This is an important difference, because the conception of the work occurs far from the equipment, and is simply treated as a tool to realise the composition. Compare this with using the computer as the composition pad to realise your ideas. The ideas are beyond acoustic or electronic timbres, they are merely the ideas which Saariaho conjures in her mind, and the limitations of the tools are the only boundaries. Compare this to working within the boundaries, within the walls of a piece of software or pad of paper.

Ending with another quote, which I believe is sometime at odds with our current compositional thinking, and more in common with our belief in scientific knowledge (not to confuse this with current meta-scientific thought which purports that we are merely travelling in circles of knowledge): “…just as the universe of Newton is contained within the universe of Einstein, contemporary music, can similarly contain, in addition to other elements, all the developed knowledge of our civilization, as well as the knowledge that we have been able to acquire about other civilizations”.