Session #2, Foundations: Musical Parameters, Mappings, and Tools

Chapter 2. Meeting 2, Foundations: Musical Parameters, Mappings, and Tools 2.1. Announcements • If you have not downloaded and installed Python and PD-Extended, please do so now • Download: most recent athenaCL http://code.google.com/p/athenacl 2.2. Overview • Events • Parameters • Containers • Instruments • Generative software tools • athenaCL and Python • Digital Audio Workstations 2.3. Musical Events • The event is the fundamental unit of music • An event can be single sample lasting 0.0000227 seconds • An event can be a note • An event can be a continuous sound encompassing a complete work • The minimum definition of an event is a start and end time 2.4. Events and Parameters • An event can be described as with one or more parameters • Parameters may be duration, pitch, amplitude, or any other collection of specifiers • Parameters may be coordinated or independent • Human musical production often coordinates parameters • Independent musical paramerers can make interest musical structures • The parameterization of musical events has been critical to the development of modern music 2.5. Event Lists • Events, defined by an array of parameters, can be collected in a list • Musical data is stored in various arrangements of event lists 2.6. Fundamental Musical Parameters • Duration and rhythm • Frequency and pitch • Amplitude and dynamics 2.7. Parameters: Duration and Rhythm • Can be measured in absolute or relative values • Absolute values: seconds, milliseconds • Relative values • Notation: quarter, sixteenth, whole • Pulse triples: (divisor, multiplier, accent) • Relative values proportional to a beat rate (tempo) • Tempi are often thought of in beats per minute (BPM) • A range of durations at different tempi [py/demo/parameterDuration.py] 2.8. Parameters: Frequency and Pitch • Pitch is a human interpretation of frequency • Pitch asserts the octave as referential unit of equivalence • An octave is 12 half steps, 8 diatonic steps (white notes on the piano), and a 2:1 frequency ratio • Numerous other distances between pitches (intervals) have names: fifths, thirds, 13ths, quarter tones • Pitch names can carry octave designation, where C4 is middle C • MIDI pitch values place C4 at 60, use 1 as a half step, and range from 0 to 127 • athenaCL pitch space values place C4 at 0 and use 1 as a half step • A range of fundamental pitches [py/demo/parameterPitch.py] • The (ideal) audible frequency range: 20 Hz (MIDI 16, E0) to 20000 Hz (MIDI 135, D#10) • Top three octaves (from 3-6k, 6-12k, 12-24k) contain spectral frequencies 2.9. Parameters: Amplitude and Dynamics • Bits: discrete digital audio amplitude levels • dB SPL: acoustic power • dBv: voltage amplitude • Unit interval spacings: between 0 and 1 • Notation: from ppp to fff • MIDI velocity values from 0 to 127 • A range of amplitude levels [py/demo/parameterAmplitude.py] 2.10. Storing Event Lists in Containers • Events can be streamed in real-time or stored in containers • Western notation (scores, MusicXML) • Musical Insturment Digital Interface (MIDI) • Open sound control (OSC) • Digital audio files 2.11. Containers: Western Notation • Events are organized around notes that specify pitch and duration • Parameter values are limited (mostly) to symbols • Parameters are isolated for instruments by staves • Parallel staves express simultaneous events • Timbral specification relies mostly on instrument assignments • MusicXML offers a standard for encoding notation • Software permits opening, editing, and playing MusicXML files • Finale and Sibelius • Finale reader http://www.finalemusic.com/Reader 2.12. Containers: MIDI • A binary representation of musical parameters • Parameter values are often 7 bit, or 128 discrete values • Parameters are isolated by numerical tags, called channels • Timbral specification relies mostly on instrument assignments (programs) • Software permits performing MIDI files • QuickTime and Windows Media Player • Virtual instruments 2.13. Containers: OSC • A hierarchical representation of musical parameters • Parameter values can be numbers or strings • Parameters are organized hierarchically with URL-like syntax • Timbral specification relies mostly on receiving device • Sending and receiving OSC data • Hardware controllers • Software controllers 2.14. Containers: Digital Audio • A micro mono-parameter representation • Store amplitude values within a dynamic range taken at a sampling rate • Signals can be mixed or stored in isolated channels • Digital audio is a timbral specification • Software permits playing and editing digital audio • QuickTime and Windows Media Player • Audacity http://audacity.sourceforge.net • Digital Audio Workstations 2.15. Synthesizers, Samplers, and Virtual Instruments • Acoustic instruments translate parameters into acoustic sound • Electronic instruments synthesize tones with oscillators or stored samples • Digital electronic instruments are built by combining basic software components • Virtual instruments are software synthesizers or samplers that respond to MIDI or OSC parameters 2.16. Digital Synthesizers • Built from combing fundamental signal generators and processors (unit generators or Ugens) • Can be designed to accept any number of initial event parameters • Can be designed to accept dynamic parameters over the course of an event 2.17. Digital Synthesis Languages: Csound • Developed in part from the first synthesis language Music 1 in 1957 (Roads 1980) • Extended and ported by Barry Vercoe at MIT (1986) • A huge library of processors and instrument models (Boulanger 2000) • A low-level language for defining instruments • A flat list of data for event lists 2.18. Digital Synthesis Languages: PureData • Over 20 years of development in synthesis, sampling, and a visual programming envrionment • Numerous related alternatives: Max/MSP, jMax, Open Sound World (OSW) • Developed by Miller Puckette, creator of the first Max (Puckette 1985, 1988, 1997, 2002) 2.19. Digital Synthesis Languages: SuperCollider • An extension of Csound archetypes into a modern language and network archetype • First released in 1996 by James McCartney (McCartney 1996; McCartney 1998) • A complete object-oriented language: create objects, manipulate, and reuse code • A server-based architecture: SynthDefs live on a server and send and receive messages and signals • Designed for real-time performance and experimentation 2.20. Virtual Instruments • Software plug-ins that can receive MIDI or OSC messages • Distributed as VST, AU, or other plug-in formats • Can employ any internal software and synthesis model 2.21. Algorithmic Composition and Generative Music Systems • May be built within a synthesis language • May be stand-alone systems • Numerous systems support multiple output formats from a single interface • athenaCL http://code.google.com/p/athenacl • AC Toolbox: Lisp based Macintosh application/environment http://www.koncon.nl/downloads/ACToolbox/ • Open Music: Lisp based visual programming language http://recherche.ircam.fr/equipes/repmus/OpenMusic/ • Common Music: Lisp based programming language http://commonmusic.sourceforge.net 2.22. A Brief History of athenaCL • Started as a way of automating the production of Csound scores in 2001 • Originally attempted to integrate a variety of post-tonal music theory tools • Gradually became a more general tool for composition • A way to test and deploy modular approaches to generating music parameters and structures • Support for output in MIDI, SuperCollider, and other formats incrementally added • Version 2 strips away post-tonal music theory tools, focuses on compositional tasks • Present alpha releases may have bugs: please report any problems to me immediately 2.23. Installing and Running athenaCL • Download the most-recent version • A distribution from Google Code http://code.google.com/p/athenacl • Via SVN command-line argument: svn checkout http://athenacl.googlecode.com/svn/trunk/ athenacl-read-only • Install in Python’s site packages • Windows: run athenaCL.exe installer • Others: extract athenaCL.tar.gz With terminal, cd to athenaCL directory Enter: python setup.py install If permissions error, try: sudo python setup.py install • Start Python • • Windows: run python.exe or IDLE.py • Others: open terminal, enter: python Start athenaCL • From within Python, enter: from athenaCL import athenacl • Others: open terminal, enter: python Enter: from athenaCL import athenacl 2.24. Running athenaCL Without Installing • Download athenaCL (as above) • Launch the file athenaCL/athenacl.py with Python 2.25. athenaCL: System Overview • Create and edit Textures (TextureInstances) and Paths (PathInstances) • Paths are static pitch collections • Textures are dynamic variable parameter event list generators • TextureModules define various approaches to create Textures • ParameterObjects are used to configure and generate parameters within Textures • EventModes define orchestras of instruments and available output formats • EventOutputs are output formats, some available with all EventModes, others available from only one • EventLists can be created, rendered, and heard 2.26. Interactive athenaCL Commands • athenaCL as an interactive command line program • Commands can be provided with space delimited arguments, or the user can be prompted for all necessary arguments • Acronyms are always accepted for arguments • cmd: view all commands • ?: get help for any command • EMo: select EventMode midiPercussion • EMi: list available instruments • TIn: create a new TextureInstance (provide name and instrument number) • ELn: create a new EventList • ELh: hear (or open) a new EventList • Commands with full arguments and sample output pi{}ti{} :: emo mp EventMode mode set to: midiPercussion. pi{}ti{} :: tin a 50 TI a created. pi{auto-highTom}ti{a} :: eln command.py: temporary file: /Volumes/xdisc/_scratch/ath2010.02.04.09.45.48.xml EventList ath2010.02.04.09.45.48 complete: /Volumes/xdisc/_scratch/ath2010.02.04.09.45.48.mid /Volumes/xdisc/_scratch/ath2010.02.04.09.45.48.xml pi{auto-highTom}ti{a} :: elh EventList hear initiated: /volumes/xdisc/_scratch/ath2010.02.04.09.48.11.mid • Setting the scratch directory to "/Volumes/xdisc/_scratch" pi{}ti{} :: apdir x /Volumes/xdisc/_scratch user fpScratchDir directory set to /volumes/xdisc/_scratch. • Editing the Texture’s temp with a WaveSine generator (space divisions matter) pi{auto-highTom}ti{a} :: tie b ws,t,10,0,40,400 TI a: parameter bpm updated. pi{auto-highTom}ti{a} :: eln; elh 2.27. Automating athenaCL Commands with Python • athenaCL command can be scripted and controlled in Python script • Permits reuse and extensions • Must create an athenaCL Interpreter object and send string commands • Creating a Python script file • Windows: use IDLE.py or another text editor • Others: use emacs, vi, or other text editor • Mac: use TextWranger (free) http://www.barebones.com/products/TextWrangler • Automating the production of one Texture: create file 02a.py and run with python [02a.py] from athenaCL.libATH import athenaObj ath = athenaObj.Interpreter() ath.cmd('emo mp') ath.cmd('tin a 45') ath.cmd('tie b ws,t,10,0,40,400') ath.cmd('eln') ath.cmd('elh') • If Python cannot find the athenaCL directory (because you were not able to do an install) you must provide to python the file path to the directory containing athenaCL import sys sys.path.append(“/path/to/dir/that/contains/athenacl”) from athenaCL.libATH import athenaObj ath = athenaObj.Interpreter() ath.cmd('emo mp') ath.cmd('tin a 45') ath.cmd('tie b ws,t,10,0,40,400') ath.cmd('eln') ath.cmd('elh') • Automating the production of three Textures [02b.py] from athenaCL.libATH import athenaObj import random ath = athenaObj.Interpreter() ath.cmd('emo mp') for x in [45, 51, 75]: ath.cmd('tin t%s %s' % (x, x)) ath.cmd('tie t %s,%s' % (random.choice(range(0,10)), random.choice(range(20,30)))) ath.cmd(“tie b ws,t,10,0,40,400”) ath.cmd('eln') ath.cmd('elh') • If you have trouble running a Python script on Windows, visit: http://www.python.org/doc/faq/windows/