Welcome to Scribd. Sign in or start your free trial to enjoy unlimited e-books, audiobooks & documents.Find out more
Download
Standard view
Full view
of .
Look up keyword
Like this
2Activity
0 of .
Results for:
No results containing your search query
P. 1
The Art of Noise

The Art of Noise

Ratings: (0)|Views: 90|Likes:
Published by Riccardo Mantelli

More info:

Published by: Riccardo Mantelli on Apr 01, 2008
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

05/26/2010

pdf

text

original

 
LinuxUser & Developer
023
 H
ackers have always been attracted byaudio, fascinated on the one hand bythe intricacies of compositional structureand on a more granular level, obsessed with howsound itself is created. Computers create an idealplatform for both areas of investigation, skilled asthey are in dissecting such structural qualities. It’shardly a coincidence that Johann Sebastian Bach’swork features so heavily in the work of pioneerssuch as Hofstadter. And ever since the ItalianFuturists in the early twentieth century embracedthe speed and violence of new technologies,making sound with machines has exerted a fascination on all manner of artists. Whencomputer hardware and languages becamesufficiently advanced in the early 60s to suit themodest needs of audio, parallel developments inaudio theory and computer science spawned anew and still relatively youthful field of artisticresearch, which now with the growth of freesoftware does seem to be undergoing arenaissance.Given that most audio or compositionalprocessing is less computationally demandingthan, say, work with video, this field does have alonger, if not richer, history. It’s commonknowledge that as the intelligent computer HALslowly dies in Kubrick’s 2001, the machineregresses through the history of its ilk, to sadlysing “Daisy, Daisy, give me your answer please,”the first song ever sung by a computer. Coded bypioneer Max Matthews at Bell Labs, coincidentallythe birthplace of UNIX, “Daisy Bell” wasgenerated using MUSIC IV, a Fortran-based musicprogramming language. MUSIC IV was obviouslyonly one iteration of the flexible language MUSIC-N. As Matthews states, “Computer performanceof music was born in 1957 when an IBM 704 inNYC played a 17 second composition on theMUSIC I program which I wrote.”MUSIC-N sowed the seeds for open sourcedlanguages such as Csound or environments suchas SuperCollider, which are still heavily used anddeveloped by artists today. What’s surprising isthat a flexible, extendible language-basedapproach to the creation of sound has beenaround since the very birth of digital audio. Andit’s possible to argue that such a methodology, away of working with art and code which stressesthe creation of new structures and environments,is inextricably tied up with the philosophy anddevelopment model of free software. Artists and computer mavericks shared ideas,code and ways of working across diverseinstitutions from Bell, MIT (Massachusetts Instituteof Technology, another landmark within thehistory of free computing), and Stanford in theUSA, to the legendary IRCAM (Institut deRecherche et Coordination Acoustic/Music) inParis. But even as early as the mid 1970s, with therelease of funky micros such as Commodore’sKIM, artists and developers could embrace a lesscentralised way of working, untied frominstitutions; a development model which stressedcommunity and openness. Indeed, the lessdemanding nature of audio computation has clear benefits, as with the rise of the low cost PC, prettymuch all but the most ambitious audio workscould easy be realised on the humble, home PC or live performance-oriented laptop. And suchuntethered environments are the perfectspawning ground for an open source approach.
 FREE INNOVATION
 Artistic digital audio is a youthful field in which themost exciting work in recent years has embraceda free software model for experimentation,production and distribution. It’s an alternativemediascape in which artists have no need of proprietary models, favouring community andexperimentation over commercial concerns. It’s allabout creativity rather than a workaday tool-basedapproach. Nearly all the apps we’ll investigateeither push some artistic or audio boundaries, or present novel ways of looking at how to work withcode and artistic material, questioning how weand machines interact and co-create. In commonwith most academically researched material,nearly all seriously experimental and extendibleaudio apps and environments have been releasedwith source code, if not strictly under a freesoftware license. The big two more traditionalmusic lab apps, Csound and CLM (Common LispMusic) are good examples of audio research toolswhich can easily be located with source on hugeacademic repositories such as Planet CCRMA(pronounced karma). Planet CCRMA, collated byStanford University Center for Computer Researchin Music and Acoustics where pioneer MaxMatthews now resides, provides a greatopportunity to witness the breadth and range of readily available open source audio apps. Therepository, which more or less amounts to acomplete GNU/Linux multimedia distro, replicatesthe research and production environment used byartist-developers at CCRMA on a daily basis. And AGNULA/DeMuDi, a tailor-made audiodistribution funded by the European Commission,is also well worth checking out (see page 80 inthis issue). There can be little doubt that freesoftware artists have no end of apps andenvironments to play with. And though newbiesmay complain that free software is troublesome toinstall and configure, Darwin and the excellentFink project have led to a burgeoning community,as OS X users join the free software fold. Thebenefits are clear and well worth repeating.Free software stresses flexibility, independenceand community providing future-proof, reusableand extendible artistic solutions for an impatient, fast-moving audio scene. Free software is all aboutthrowing away the rulebook; after all computer science and digital audio should have noconstraints at such a youthful age. In recent years,as the scene has snowballed, ad-hoc groupings of artists based around specific languages,environments or approaches have quickly formed.Grass roots organisations such as Bek in Norway, V2_labs in Holland and Bootlab in Berlin havesolidified such groups, providing resources andsupport for artists and developers. And althoughPd (Pure Data) and SuperCollider have probablythe widest followings, plenty of artists are codingtheir own apps in a range of languages. Freesoftware audio artists share a healthy range of geek interests and a passion for playing live,pushing the envelope well past a lone guy withlaptop scenario to embrace all manner of noise
The Art of NoiseThe Art of Noise
022
LinuxUser & Developer
The ArtOf Noise 
In the second of a threepart article examining thegrowing artistic uses offree software, Martin Howseprobes the world of artist-developers working primarilywith self-coded audio apps
Networked performance system for junked PCs. ap atElectrohype2002The aftermath of an ap performance using both ap02and gdapp software in Malmo, Sweden as part ofElectrohype2002sound and image as raw data re-processed live by theself-recoding ap02 application
 Hacking the global soundscape 
 
024
LinuxUser & DeveloperLinuxUser & Developer
025
producing devices and computationalmachinery. With ap’s computer controlledrecord decks and neural synths, MumbaiStreaming Attack’s recent TRAMJAM sequencingsamples using 32 trams in Vienna, and codingpoetry slams in Denmark, the free softwareaudio scene obviously has little in common withthe dull Cubase clones and tired software synthsproduced by proprietary models. The history of computer audio is all about free innovation.
MAXED OUT
Hacker-artist’s favourite environment, Pd, reallydoes have roots stretching far back into the historyof computational audio creativity. Based aroundthe Max, and later Max/MSP, extendible visualscripting environment for audio creation, andauthored by the very same Miller Puckette, Pdtakes up where the proprietary Max/MSP left off,thanks to an incredibly lively developer community and free software model. For usersunfamiliar with the Max/MSP or Pd environments,this visual model can easily be described asallowing the interconnection of various boxeswhich can pass both data and control messages. As you’d imagine from its very name, Pure Data,the flexibility of Pd lies in treating audio solely asdata which can be manipulated, mixed andgenerally thrown around with a huge number of maths, control and generator objects able tochange data pathways, structures and events. With such a flattening of audio as simply raw data,the possibilities are endless, with additionalextensions also able to manipulate data asgraphical content or even through advancedneural net architectures.Collections of connected boxes are calledpatches, and these can quite easily hide further sub-patches which simply pass data inboth directions to the main patch. Patches arecomposed of connections between object boxes,messages, GUI boxes such as VU meters or slidersand comment boxes for documentation purposes.It’s a reasonably hierarchical approach but mostartist’s do adapt Pd to suit their ownmethodologies, either through extensions whichcan, say, add language scripting possibilities, or simply through the abstractions provided bysub-patching.
GO FORTH AND MULTIPLY
The easiest way to understand Pd is quite simplyto get up and running, and begin to play with thehuge range of objects available. Under mostGNU/Linux distros, Pd is trivial to install, with fewdependencies and uncomplicated configuration.It may well be worth messing with the low latencypossibilities in recent kernels, or patched earlier kernels, and Pd does play well with ALSA andJACK, but a base setup is easy to achieve withreadily available walk-through tutorials. Pd issuitably platform agnostic, also running under both Windows and OS X. Documentation isintelligently embedded within the Pd package,and commented Pd patches are easily accessible from within Pd. Examples range from simpleoscillators through more complex control andaudio patches, to advanced Fast Fourier Transform(FFT) examples, converting data from the timedomain to the frequency domain for further analysis and manipulation. Reference patches arealso included which document every single objectwithin the base Pd package from sine and cosinegeneration objects, through control, network andpiping objects, to MIDI and sampling objects.Simple patches may use only a small range of these, but it is easily possible as we’ll see to createseriously spidery patches which pack in hardcorenetworked and control functionality. And with ad-hoc groupings of Pd-heads sharingideas and code, extensions have grown up aroundPd which allow less technical users to more simplycoerce Pd towards their own particular artisticvision. Tom Schouten and Yves Degoyon,occasionally working together live as BULT, areprime movers on the Pd coding scene, betweenthem producing both graphical extensions for Pd,and a huge number of intriguing externals. Onthe audio side these include Icecast andSHOUTcast externals, vocoders, compressors andsome seriously complex math and DSP objects.Schouten has also recently released the highlyadventurous packet forth (pf) package, which canbe used both with and without Pd and Schouten’sPDP multimedia extension. Pf throws together theForth language, Pd, Lisp, Unix scripting andhardcore DSP and it could easily be considered asa multimedia glue language which can be usedwith a variety of interfaces. Schouten has recentlyprototyped an inferior pf interpreter process under GNU Emacs, which also opens up some seriouslive coding possibilities.
PARTICLE ACCELERATION
 At first glance the popular SuperCollider environment does appear to offer a radicallydifferent approach to audio work than Pd, thougha good few Pd-heads, such as Farmers Manual, dodivide time between them and both apps doshare common roots and an environmentalcoding approach to audio. SuperCollider, or SC,can easily be viewed as a textual code-basedenvironment, in contrast to Pd’s obviouslygraphical model. Both approaches haveadvantages and disadvantages, particularly whenit comes to live work, but it’s easy to see that Pdand SC could quite easily compliment each other.Indeed, SuperCollider can quite simply beinterfaced with Pd, to provide the best of bothworlds. Pd appeals to the artist who wants throwtogether, to jam with, disparate data sources andcontrol, maybe interfacing complex hardware,whereas SC is perhaps more geared towardsgenerating intricate compositional structures.SuperCollider has a long history centeringprimarily on the work of one James McCartney,previous to its now open sourced SC3 iteration.It’s interesting to see our old friend Max pop uponce more, with SC1 morphing out of McCartney’s Max object Pyrite, though SC itself ismore about MSP style DSP work than pure Maxcontrol. SC also owes a good deal to the evenmore venerable MUSIC-N apps. It was MUSIC IIIwhich introduced the concept of a unit generator,a subroutine that would create a specific kind of sound, which is so central to SC. The programconsists of two applications, a client which is thelanguage itself (sclang) and a server, scsynth,which handles the DSP work and sound synthesis. With a highly flexible programming languageengine, and supremely networkable architecture,again making use of OSC (see below), it’s not tootough to see how far you could push SC. Sclanghas morphed considerably through variousversions, borrowing differing features, syntax andmodels from languages such as Smalltalk,Scheme, C and even J along the way, but the coreof math operators, oscillators, noise generators, filters, controls, delays, samples, event spawningand I/O wrapped up in an extreme objectoriented model with messages and classes remainsas the SC approach. Unit generators, such asdiverse objects which process or generate sound,are an essential part of SC, and the controlngenerator unit class, for example, allows for external control sources, graphical sliders or programmatic processes to control parameters.Examples would include MouseX and MouseYclasses. Plentiful online tutorials, wikis,documentation and a fresh, active community areon hand to provide examples and support,though at first SC may well seem like a toughcookie to crack, with a code-based approachseemingly at odds with most artists’ methodology.Getting up and running with SC under GNU/Linux may also present some problems for the lessexperienced user, though quite thoroughdocumentation is available. JACK and ALSA areabsolute necessities, and low latency audio ispreferable. With these in place, SC3 can easily bechecked out from CVS and compiled. Somewhatlengthy configuration of elements such as JACKinputs is necessary before scsynth is up andrunning, and sclang itself does need a startup file for class and environment variables, but all this iswell documented. Online material is also available for those wishing to extend SC with plugins, andthe intriguing architecture of SC should make of this an interesting exercise for the artist-hacker.
 As the intelligent computer HAL slowly dies in Kubricks2001, the machine regresses through the history of its ilk, to sadly sing Daisy, Daisy,give me your answer please, the first song ever sung by a computer
The Art of NoiseThe Art of Noise
Alex McLean rises to the live coding challenge during the intriguingLondon Placard headphone festivalPd-heads debug the latest spidery TRAMJAM patchwhich also uses Python to tie together a supremelycomplex networked and hardware system

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->