Re: Choreography vs. Composition (and environments)

Darren Kelly (
Thu, 8 Jan 1998 15:20:27 +0100 (MET)

> > How does a dancer exercise control points ? There are
> > ONLY signals and filters (with settings). That's all there is:
> >
> >
> > Here are my suggestions:
> >
> > input signals = "performance data"
> > (generated by d(r)ancers)
> >
> > filter settings = "signal-to-sense algorithm"
> > (chosen by musical/lighting/stage composers/mixers)
> OK, but I think the term "filter" is misleading - it suggests that
> information is being removed or rarified, rather than used in a
> generative manner.

No, filter is exactly the right term. The task being performed here has
_NOTHING_ to do with dance and music, it is a very general signal
processing task that has been performed for over one hundred years,
and for which mathematicians and engineers already have language. You
have input signals and filters. I have discussed my gestural MIDI control
work with signal processing engineers here at DESY and elsewhere, and
the terms input, signals, filters, and output are completely accurate and
completely adequate. They at most need to be supplemented by some
dance-, music-, and light-specific terms.

A filter does not dilute or rarify anything unless you want it to. You are
perhaps thinking of optical filters in photography which can (but don't
always) reduce the signal strength substantially. This is a very specific
case, and atypical.

As an alternative, think of a traditional mixing board. The signal
comes in at one end and can be amplified (gain), "filtered" by pan knobs,
"filtered" by EQ knobs, etc. The signal can also be sent to other filters
(FX) then recombined. This is essentially how DranceWare++ is designed
[G'day rowntree3d].

> But if we stick with "filter": presumably a
> composer/musician could be improvising with the filters' processing
> algorithms at the same time as the dancer is producing the initial
> signals.

Exactly, unless you wish to control the programme with one hand and wave
the body around standing at the heyboard. Not ideal, I've tried it.

I'm considering using the sensors attached to the body as a type of mouse
so the performer can control things remotely. There are a number of people
working on alternatives to mice, including Don Buchla (maker of Lightning
II). To some of you this idea will be new, but to virtual reality people
it's old hat.

> The signal-to-sense algorithm could also be regarded as dynamic,
> generated or controlled by a musical collaborator during the dance
> performance. This is the sort of interaction which interests me.

This is the sort of interface I'm working on. Exactly that.


Darren Kelly | | |
| \ 0 / |
DESY -MPY- (Deutsches Elektronen-Synchrotron) | * o |
Notkestr.85 | - D E S Y - |
22603 Hamburg | o * |
Germany | / O \ |
| | |
phone: +49-40-8998-4569 | |
fax: +49-40-8998-4305 <--Note change | |
e-mail: | |
Amandastr.40a, | HOME |
20357 Hamburg | zu Hause |
phone: +49-40-4322602 | The PLAYhouse |