il duello

– yakanama

i always find it interesting that the majority of those who ‘create’ dance technologies are unable to deal with criticism of their work. This is especially true of the programmers, who for the most part don’t have a clue about dance, or dance studies … tony schultz is a perfect example of this.

tony posted a belated comment on my post chromatic particles in which he takes offense to my critique of his work. In spite of tony’s protestations, his work is flawed for the application he suggested it might be used for … documentation and notation. go back to doc&rec 101 tony, and take some notation classes while you are at it.

the model does not get “confused” or fail to “work properly” when the legs fuse together, it simply represents the legs as a single particle when there is occlusion between mirror limbs. this single particle representation of the legs together is sufficient for the detection algorithm to function and for the viewer to perceive what is happening, and thats all the only work i need to get done.

And just for reference from his first post …

I primarily talked about my thesis research and how it can be applied to the problems of notating and documenting dance. My research uses computer vision algorithms to represent the body as a set of chromatic particles. Once the body is reduced to numbers it becomes possible to automatically recognize different poses. Once these landmarks are identified the computer can generate a map of the movement space in the form of a dance graph.

The bottom line is that your approach is deeply flawed when applied to dance notation and documentation.

If your software deals with occlusions or mirror limbs by blending them into a single limbs then it is failing to identify landmarks correctly. So yes, your software does get confused when trying to map the landmarks, occlusion is not equal to fusion, it is misidentification. so is not a useful contemporary dance notation because it fails to notate what is happening. anyway splitting hairs over the function of the markov models and the aims of your software is does not distract from the overall problem. Maybe you also don’t realize that orientation is a component of posture, and the visual representation you are using is very poor at delineating between different orientations.

You cannot reconstruct properly from the software / example you have shown … Representation of two legs as a single object does not tell me what is happening with the legs. All i have seen of your work is automated systems that replicate and (re)visualise post-modern (post-judson) choreographic principles and practice … Its no wonder that sara rudner and twyla tharp find your work interesting. I mean really, they are hardly guiding lights in dance and technologies practice. telling me they find it interesting does not change the fact that the chromatic particles are a poor tool for notation and documentation.

You don’t really know what to look for in the dancing body to select and apply the appropriate mathematical models.

And yes, i have read your other posts, which equally show a lack of depth in dance and dance technologies history / theory. You seem deeply unaware of what you are replicating in dance technology and the finer aspects of dance arts. Don’t confuse me for a twirly that does not understand your mathematical / computational approach … I do. And i can see where it is misapplied. So if you want to really defend your work, share your code.

i am not sure what you mean when you refer to my style of choreography. the video presents only a single path derived from the graph extracted from a set of improvisational data.

Firstly your explanation of a single path here is poor, as arguably each particle has a path. So i will assume you mean the general motion path (which is a crude approximation, like your representation). The graph, or whatever you want to call it has stylistics elements common to hiphop/street dance vocabulary. thats what i was referring to as your choreographic style.

Its also interesting you don’t show the original data set and source, was there to much divergence between the two?

ti 4 n00bs

i had a few emails demanding a better and simpler explanation of temporal interference (ti). apparently my description was not very n00b friendly or technically accurate enough for some people. so whilst i’m not your dancing bear,  here you go …

• dancer moves
• dancers movement triggers theremin
• theremin generates two analog signals (one from each antenna)
• analog signals are converted into midi data (two midi signals)
• midi data (values 0 – 127) sent to max/msp/ jitter
• max utilizes midi data in multiple max patches for control and generation of audio/video, e.g.:
  • zoom
  • saturation
  • stills & video capture
  • audio samples
  • pitch (frequency)
  • perceived volume (oscillation)

… and thats the way it is.

inside the interface

– *aletheuo*

i already gave you an overview of temporal interference (ti), so let me share with you how i engage with interface works, and a little of my experience playing with the ti installation.

mapping: i always begin my mapping out the interface, transcriptions (how an input is re-written to a new output) and the choreographic / conceptual ideas. this usually requires asking lots if different people questions, especially if multiple choreographers and programmers are working on a single project. i make notes at the end of the day for reference using yed, with the generic java version (on my usb stick) so i’m fully cross platform (windows, osx, unix / linux) which is useful given that range of operating systems i’m exposed to. once i have the map worked out, i commit it to memory.

texturising: once i have the map, i texturise it in my body and mind, the textures have patterns which reveal the connections to other locations. essentially it is a textured and colored perceptual score so objects that are blue might all draw on data from a specfic sensor (e.g. the pitch antenna). i visualize the map as a room, similar in structure to the installation and a costume for me which is vastly different form what i will actually be wearing. in the performance i shift between the map, and the real physical context. as most installations are not truly interactive, interpret specific system outputs (on a performance by performance basis) as tuning score calls (play, pause, reverse, shift, replace, open-close, reduce, mirror, speed up-slow down, end). so a particular visual event (green flashing) might cause me to repeat the last 5 secs of material.

with ti, i was able to deduce the basic mapping of the system drawing on my prior experience of similar works. the transcriptions are quite simple and seemed to be basic max-msp patches, but i kind of liked that. the simplicity made it much simpler to map and texture. what i found slightly frustrating (as i noted before) is that the sensitivity and replicability of action in the interface were very limited. when watching the performance there were moments when the dancers seemed to be equally frustrated in their attempts to trigger the interface. the performance work itself was highly choreographed, which is not a problem except that i didn’t feel the choreographer or dancers really understood what they were working with.

i think its when artists don’t understand the performance systems they used that the old conflict between technology and humans issue arises. humans are at one with tools and technology, our development and evolution over time has been dependent on tools, and exploited tools. technologies are integral to human life, be the low-tech (fire) or high tech (silicon chips).

doug posted to say he will interview the creative team behind ti, i’d also like to hear from one of the dancers, then he would have a girl to interview!

if you want a basic explanation of ti try ti 4 n00bs.

note to self

– rose&olive

when asked by the choreographer what your thoughts on the work are, remember to simply repeat what they have said to you. off the cuff crit’s 30mins before the show are (usually) not welcome … look pretty androgynous and dance.

actually i should learn not to say whats on my mind in the presence of people more often. i tend to respond to the question rather than realizing its an attempt to pass time. as a dancer people generally want to see your interpretation of a work, not hear it. in spite of all that thinking dancer guff *, dancers who conceptually engage and articulate their thoughts are considered suspect. its as if the dance can’t be in your body if you establish a clear critical (and theoretical) position on it.

just think about it, how often do you read in depth article from dancers about how they engage with, interpret and recreate dance works.

* all dancers have to think, what they really mean is critically informed dancers.

missing a particular friend of mine, wish i could do a lindenberg and see them,

the winds of march that make my heart a dancer
a telephone that rings but who’s to answer?
oh, how the ghost of you clings
these foolish things remind me of you

notes on temporal interference

last week i went to see temporal interference (ti), the next day i chatted with matt about the performance technologies in the show before i flew out to toronto. i’m not able to separate out our conversation so i’m just acknowledging matt’s input.

the primary interface for ti is a theremin suspended from the ceiling from which the pitch and volume antenna are dressed with plastic tubing that extends to the floor. an iSight (webcam) pointed between the two antenna towards the audience is the secondary secondary interface controlled by the theremin and max/msp . the installation also makes use of a sound system, video projector and scrim which you can see in the picture above and my map of the performance space.

you have probably heard a theremin even if you didn’t know what it was. the theremin’s two antenna have a electromagnetic field in which the resistance of your body actives the theremin. For ti the theremin used a ‘soft’ set-up to have a large electromagnetic field with minimal background interference. this type of set-up results in a significant loss of fine control and sensitivity in the interface. a ‘hard’ set-up is more complex and involves modifying the physical components of the theremin for greater range and control. hard set-up’s are complex and i believe that this was the desired set-up for ti, but was sadly not achieved.

From my brief experimentation (the system was live before the performance) it seemed as is the theremin had non-linear control. Given that the soft set-up used limits the control with which the theremin can be ‘played’, linear control would heave been more useful. Its much easier to ‘learn’ how to play a theremin with linear control. the accuracy with which you can play a linear control theremin is amazing. a few years ago i worked with four theremin installation and each thermin was expertly set-up. i really enjoyed developing the performance over time as the interface was regular, even if the audio processing shifted.

 

• Linear control: the physical distance between intervals /switches is evenly spaced along the entire interface (like a piano).
• Non-linear control: the physical distance between intervals / switches decreases along the interface (like a guitar).

i can usually tell if the interface i’m working with has linear control by watching the set-up. if the tech people are using a specific (measured) calibration method then it normally means there is linear control. this does not happen as much as i would like. interface control is important because it is how the performers (dancers or public) interact with the installation. this is a rough sketch of the interface & processing mapping in ti (thanks matt),

the midi values from the pitch and volume antenna are used by various max patches to affect audio (pitch, oscillation, synthesis etc.) and video ( zoom, fps, saturation, image permanence). in another post i might talk more about my physical experience of playing in the installation, but for now i suggest that you go play with ti and watch the performance yourself this weekend.

you might want to read ti 4 n00bs if you want a simpler version of this post.