Nov. 25, 2007
Here in the European heart of global finance that is London, there are no shortage of closed, corporate-sponsored, invite-only events promoting proprietary products with secret formulas and patented ideas. A refreshing break from this trend is the perennial [tag]Openlab[/tag] series of events, now in its fourth incarnation. Openlab is a loose collective of artists, computer industry professionals, and performers whose main goal is to spread free software an ideas through events with talks, workshops, and performances.
Openlab4 took place 25 November at Melange, an aspiring art-friendly venue just north of east London’s art-clogged arteries of Old Street and Shoreditch. Coincidentally, this date was two days after the American holiday of Thanksgiving, where Americans everywhere celebrate the legendary day when the native Americans instructed the foreign Pilgrims in the technology of producing and preparing American food, especially wild turkey.
In contrast to last year’s Openlab3, which took place in a reclaimed art space in Peckham and functioned mainly as an Openlab-curated art exhibition space and performance venue, this year’s Openlab4 split its time evenly between talks and workshops, and nighttime performances. Since the members of Openlab are veteran presenters on the lecture circuit, having given workshops and presentations at conferences too numerous to mention, from Piksel to Haskell, the caliber of the talks was very high. There was no”winging-it,” unless you define that as showing up with a well-honed, bullet-pointed presentation and a collection of videos on topic. The performances were equally excellent and battle-tested, proving that you don’t need to shell out exorbitant amounts of dosh in order to put on a good show.
Still, if anyone stole the show it was Jagannathan with high-energy presentation on din, or “digital instrument.” Clearly a labor of love, din is also a marvel of what one person can create with enough skill, dedication and bezier curves. Jag’s energy bubbled over into his presentation, where he just couldn’t bring himself to wrap-up despite repeated hand-signals from co-organizer Chun Lee. Fortunately, I’m sure that the 50+ members of the packed Melange basement audience forgave him for running late.
This past summer I had the chance to see an early version of din at Dorkcamp, and while that was an impressive translation of traditional Indian instruments and their pitch-bending performance properties into a simple, mouse-controlled program, it was nothing like the multifaceted din unveiled at Openlab4. The din at Openlab4 had a variety of different looks to it. Sometimes, notes were buttons, plucked like sitar strings by the mouse cursor; other times the sound was a continuous drone controlled by the mouse scrolling left-right and up-down as if on a fretboard; yet other times sound was a combination of buttons and bezier-curves and pliable as a gooey dough.
The “Why Doesn’t Anyone Use Bezier Curves in Music?” portion of his talk had us sold, though. Inspired in part by Bob Lang and Nick Collins’s SplineSynth, Jag explains:
“… I am surprised that bezier curves are not used in the mainstream applcations (Sorry I exclude university research published in closed journals I can’t freely access) to control almost all aspects of computer sound. One reason given was people like to build sound complexity by hand rolling the number and amount of harmonics. Bezier waveform does the opposite, i.e. you can’t control what harmonics are there and how much of each is present. So there is a loss of control.
However with the version of din I showed last we can build sound complexity by adding up the contribution of resonators driven by bezier waveforms. A library/palette of bezier waveforms with FFT done on them can be made available that the user can combine and hear what it sounds like. But then it also depends on the goal of the artist/project too. Sometimes you are trying to invesitage new timbres and monkeying about with the bezier waveform is a quick way to sort of ‘browse’ the timbres possible. If we find an interesting timbre, we could do a FFT on the bezier and see the harmonic distribution. That might inform the traditional way we build up computer sound too.
However, using bezier waveform for envelopes is very powerful. This is because with just one Catmull-Rom spline (which is just a bunch of beziers linked together) we can do ADSR or any combination like say ASDSASR and apply it to any parameter. The channels editor in the last version of din is an envelope editor for both channels ie beziers were used to do AM on the waveform.”
Alex McLean started the night off with Vocable Synthesis, a project for that investigates how a computer-based synthesizer can use actual “words” to create sounds. Specifically, it allows a musician to make sounds by inventing and typing onomatopoeic words as part of polymetric rhythms. He has a video of it in action on his website which makes the concept a bit clearer. Aside from the technical aspects of building an audio synthesizer that works off the timing and “sounds” of typed words, Alex’s talk about the inspiration and history of words-cum-sounds was equally as interesting.
Ed Kelly presented his Meta Studio, a collection of feature-packed modules for making sounds and music in PureData. Ed created Meta Studio out of a frustration with having to start from scratch each time he set out to make a new sound composition. Besides including the basic step sequencers, sound generators, and looping tools that fill out most commercial audio applications, Meta Studio includes some ingenious ways of giving the sounds more of an analogue feel. Also included are some new concepts such as tabmask~, an FFT phaser that creates a mix of two sounds’ component frequencies.
Claude Heiland-Allen unveiled the mysterious GraphGrow, which is a system for creating and rendering complex, recursive maths like fractals. It’s a bit similar to Electric Sheep, except designed to be less organic and curvy and more hard edges and points. For further information I will defer to Claude’s online presentation, which is very well organized and clear. Also make sure to check out his collection of videos generated with graphgrow.
Both Rob Munro and Chun Lee introduced new OSC-based trackers, which are types of sequencers. Chun’s tracker runs in emacs (emacs!) while Rob’s is a Java application in itself with lots of features and a nice timeline GUI. Both send out OSC (open sound control) messages and are designed to be used to control other software, such as PureData (Pd).
Robert Atwood presented Jackbytes, an opensource tool for getting FFT (Fast-Fourier-Transform) data into Processing and other applications that work off the realtime analysis of music. Openlab and Cracktux member Oli (yesyesnono.co.uk) has been using jackbytes to create intricate visual performances for about a year, now.
Andy Farnell gave his gold-star “Game Audio Design In PD” presentation, which gets better and better with every presentation. A lot of it is available online, and well worth a look if you have any interest whatsoever in sound. From simple sine wave oscillators to banging on metal object to modeling bird sounds and boiling kettles, this is the most complete tutorial to sound design that you can find. Andy is also working on a textbook in this area, which is welcome news to us all.
Evan Raskob had the lone workshop of the day, bringing together the unsuspecting audience into a participatory “human live-coding” session. Participants became pieces of a human-powered synthesizer by donning signs that explained their role (“metronome,” “counter,” “clap”) and then were connected up via bungee cords (“patch cords”). More people decided to take part than expected in this experimental exercise in teaching computer programming, creating a bit of chaos, but that only added to the event’s analogue charm.
Pixelpusher + KRGN + Robert Atwood