Motion Bank: Two

Re-imagining Choreographic Ideas

A Few Questions and Responses

Posted on November 27th, 2013

Meta Academy compilation of images

Meta Academy compilation of images

On Tuesday, I had the pleasure of talking with Nik Haffner, Marlon Barrios-Solano and their students at the Inter-University for Dance Berlin (HZT Berlin) about our TWO project and the upcoming Motion Bank launch in Frankfurt. The students talked about the connection between dance and technology and asked how I came to work in this field of research. It reminded me how little I think of all this as “technology” and how much I orient around the human collaboration aspects of our work. It is the relationship between computing ideas and choreographic ideas that so defines my research in the last couple of decades. And it is the engagement between the culture of art and the culture of science that has become our true methodological space. Interdisciplinary work is intercultural work.

Here are a few more of the questions from the students and my responses:


What is the TWO project? Is it two different scores for two artists?
The TWO project is ONE digital score focusing on the thinking body and dancing mind (a phrase I am borrowing with permission from the illustrious David Gere and his Introduction to Taken by Surprise, an excellent compilation of articles on dance improvisation). So it is one project. One Score. Two artists. In some writing I did for the project I recently explained it like this:

Our work begins and ends with two dance companies. Unrelated to each other. One from the US (the Bebe Miller Company) and one from Europe (Thomas Hauert’s Zoo Company). Both are currently choreographing improvisation for performance. And both are engaging directly with the nature of human consciousness. When we watch the dancers, we are watching them at work. We are witness to the concentration and forms of attention that they bring to the moment. And we are witness to their habits, tendencies, attention, impulses, and memories in action. In this project, we have selected two working strategies each from the two companies to shed light on and bring us into a direct encounter with what the dancing mind and the thinking body. In an early storyboard for the project I called it The Dance of Attention but that was actually too limiting.


How were the artists for Motion Bank chosen?

For our part it was very intuitive really. We wanted to start from a different place then we had started with Synchronous Objects. We decided to start from a choreographic phenomenon and then look into the processes of two different artists to see what insights their ways of working might shed on that phenomenon. We were looking for people we could enjoy working with who also had a relationship to deep process and had some small-scale works (2 or 3 dancers). Bebe Miller was a natural fit because she was engaged in a duet project with her company and we knew a lot about her working methods. We found Thomas Hauert’s work through our friends at Germany’s hidden gem, the PACT Zollverein Center for Choreographic Research. PACT helped support our work by funding a residency for me and Thomas to exchange ideas and everything just unfolded from there. I immediately appreciated Thomas’ sense of processes that generate movement and his interest in the cognitive challenges that improvisation can involve.


Did you make everything you planned to make? Will you share the data so other things can be made from your resources?

Deadlines always mean something gets cut and that’s certainly the case with us. At some point you have to “put the show on the stage and turn on the lights.” We hope to add a few more animations and graphs in the coming months but what is online is a full representation of the project. We would love to share the data from this project and Synchronous Objects any time there is interest. Motion Bank is a kind of open source initiative. When Forsythe dreamed it up he really wanted it to act as a catalyst for other artists to explore digital scores and traces of choreographic thinking and to make tools that can be used by all kinds of artists for even more projects. I think the Piecemaker tool they created is super useful and the publishing system for the scores. And I hope the ideas are too.


Can you share more about the scores and ideas you selected from Bebe Miller and Thomas Hauert to discuss perception/cognition?

While the two dance companies we are focusing on are both interested in the forms of perception and consciousness accessed and enacted in improvisation, they explore those questions with very different strategies. We focus on four of those strategies in the sets you’ll see in Motion Bank at the launch on Thursday. The project is divided into four sets with the following titles: Habit, Tendency, Impulse, and Memory. I look forward to having it out in the world and knowing what ideas, questions, critiques, interests it brings.


Screen shot from our Motion Bank project TWO that will be online November 28th


Preview: Thomas Hauert, Impulse Set

Posted on November 27th, 2013

Impulses are starting points initiated without a plan being made on how the entire movement is going to develop. In this conception of movement, you don’t have an intellectual image of a whole sequence of motion but you expose your body to impulses, you vary and accumulate rules, initiations, directions, connections, tensions, release, and let the solutions emerge. —-Thomas Hauert

From left, Mat Voortner, Sara Ludi and Thomas Hauert demonstrating one external impulse strategy for priming the dancer's attention

From left, Mat Voorter, Sara Ludi and Thomas Hauert demonstrating one external impulse strategy for priming the dancer's attention

Interactive Attentive Agent, an algorithmic metaphor for Pressure Assisted Solo strategy devised by Thomas Hauert

The body keeps finding solutions to the most complex and unpredictable cocktails of forces and directions imposed on it is exhilarating and liberating way. —- Thomas Hauert

Preview: Bebe Miller, Tendency Set

Posted on November 27th, 2013

We are generating movement strategies that unfold in ways that are not as feasible in step-by-step choreography, for us. We specify the intentional body-mind-set and allow the movement articulation – the dancing – to respond to the frame or strategy.” —- Bebe Miller



The Motion Bank interface is set up in “Sets” that are devised by the creative teams working with the choreographers.  Motion Bank hopes that in the future, visitors to the site will be able to make their own preferred sets of related information. The four sets that we have curated for this premiere on Thursday November 28th focus on a handful of the mind/body strategies that these makers use in creating performance improvisation. For example, Bebe Miller has a longstanding interest in the movement tendencies of her dancers and how these can be distilled to create states for improvisation. In our TENDENCY set, we have several different windows into three states that Bebe has devised over the years, Risky Weight, Story State, and Drive State. States are the other side of HABIT and that is the focus one of Thomas Hauert’s sets.


I find myself drawn to the pull of attention, the sweep of action, the arrested moment, the pressure between people.” —-Bebe Miller



Preview: Bebe Miller Company, Memory Set

Posted on November 24th, 2013

I am very interested in the layers of remembered events we pass through daily.  Humans are built on memory.” —– Bebe Miller

Image from Video of Verge Redux, January 2013. (Pictured from Left: Bebe Miller, Angie Hauser, Darrell Jones)

As we work to complete our TWO project and prepare the assets for publication in the interface developed by Motion Bank, I thought it would be nice to post a few images and quotes from choreographers Thomas Hauert and Bebe Miller. In a series of posts starting today I will sample what we will premiere next week.


The Motion Bank interface is set up in “sets” that are devised by the makers but they also hope that in the future, visitors to the site will be able to make their own preferred sets of related information. The four sets that we have curated for this premiere on Thursday November 28th focus on a handful of the mind/body strategies that these makers use in creating performance improvisation. For example, Bebe Miller has a longstanding interest in memory and the histories we carry in our bodies. One strategy she and her company have devised over the years is a way of working they call Redux.


In our Memory Set for Motion Bank, we have video, textual scores, information graphics, and two interactive games that bring you into an encounter with Miller’s work and the possibilities of memory as rich resource for creativity.

Screenshot from Redux Interactive

There is an aspect of improvisation that asks you to shuffle through the various layers of right now in order to respond directly to the situation. We as a company have invested many years in researching this relationship between memory and the dynamics of body/mind/place/time. Redux is one example of this. ” —– Bebe Miller

Risky Weight in Bebe Miller’s work

Posted on July 12th, 2013

One of the ongoing interests in our project is Bebe Miller Company’s use of what they call Risky Weight. In developing our terminology, Norah and graduate research assistant Malory Spicer have been compiling descriptions of Risky Weight from the dancers, Bebe, Talvin Wilks the dramaturge, and our consulting Laban Movement Analyst Melanie Bales. When reading these descriptions I was drawn to this definition of Risky Weight: the whole weight is a little more at risk and less linear and sequential to the body.


As we were working on this concept, Norah asked me about the possibility of tracking center of gravity in motion capture data and adding this information to our understanding of Risky Weight, a concept that has to do both with the viewer’s perception and the dancer’s use of his or her weight and gravity. To open this description further I assumed that this means creating situations when a performer’s body gets close to losing its balance, subsequently gaining stability.


In researching center (and line) of gravity and base of support in relationship to risky weight I discovered several perspectives on the estimation and perception of body balance. The precise biomechanics based estimation requires knowing the body weight and considering percentage of that weight associated with particular part of a body (see calculations detail in this ASU Biomechanics explanation).  The visual perception perspective, based on the qualitative approach to this problem in animation is reflected in the collection of tutorials at the Animation Physics website.  It demonstrates that a less precise approach to estimating center of gravity is still valid as long as the main principle is followed:

The body is off balance when the Center of Gravity (CG) is outside the Base of Support (BS). Thus the most stable position is when the two overlap (ie the person is laying flat on the floor), the next best thing is when the CG is directly above the BS.


Since Motionbuilder (our motion capture processing software) provides a basic operator which can estimate the CG (not as precise as biomechanics approach, but perhaps close enough), I used it to visualize the relationship of Center of Gravity, Line of Gravity (vertical line passing through the CG into the ground) and Base of Support. This exploration in Bebe’s work seems useful since the risk the performer takes is that of falling, which happens when their body is off balance.


In order to establish the BS I connected each foot’s ball and heel joints of the mocap skeleton into a rectangle stretched between the two feet.  The BS is reduced to a triangle when one of the joints is off the ground and to the area of one foot when the other foot is lifted. My threshold for when the foot is considered to be off the ground is set to 20cm to avoid excessive flickering caused by frequent shape change for the base of support, this can certainly be adjusted for higher precision if necessary.


This quick study may reveal that the qualitatively observed risky weight aspect of Darrell’s performance probably does not come from the state of physical imbalance. The occurrences and durations of this state are not that significant. Again, I am inclined to hypothesize that the viewer’s anticipation of the body becoming off balance is what creates the impression of Risky Weight but there is more to be explored here.


– Vita Berezina-Blackburn, Animation Specialist/ACCAD

Momentum in Bebe Miller’s work

Posted on July 8th, 2013

In examining the motion capture work from our January residency with Bebe Miller Company we are becoming interested in the possibility of tracing the momentum transfer through the body. The paths of momentum transfer reveal patterns of movement sequencing in the body. In both Darrell’s and Angie’s performances it is evident that several motions often happen concurrently. Performative tendencies that focus on simultaneity and dis-coordination of body segment motions are likely to interrupt viewer’s ability to anticipate movement and thus create fresh and surprising performance qualities.


There are a number of ways to approach visual representation of movement sequencing, and I thought it might be interesting to see this by making visible the sequences of joint rotations.  To set up this visualization I hypothesized that the initiating joint would rotate first, sending the waves of sequential rotations to the other parts of the body.  In the video of Darrell’s performance it is easy to notice many instances when sequences of joint rotations are initiated simultaneously or with minor offsets in time . I have used color coding to present two types of joint movement analysis.  The red figure indicates angular/rotation speed of joints (A) (the brighter the red the higher the speed), the green figure indicates linear speed of joints’ locomotion (L). As opposed to joint rotation speed, visualization of the linear (joint translation through space) shows a more even speed value distribution that is more sequential.


Since these may be a bit hard to notice at high playback speed of the captured motion this movie is at 1/4th speed.


While working on ideas about momentum I ran across a video on momentum transfer in parkour.   In this video, the second jump comes as a surprise. In order to get additional momentum to make the second jump possible, the landing position after the first jump had to be exaggerated… At a quick glance the shape of the body at the time of the first landing seemed to be that of trying to slow down.  By exaggerating this pose through leaning back and further lifting the arms backwards, it is possible to gain additional momentum. This hidden quality of getting the additional momentum is very intriguing as it also serves as a way of hiding anticipation of subsequent action. This triggers parallels between the animation concepts of motion anticipation and exaggeration and Bebe Miller’s ideas of “interrupting the inevitable” as well as “furthering”.


– Vita Berezina-Blackburn, Animation Specialist/ACCAD

Establishing Vocabularies for Better Collaborative Work

Posted on June 19th, 2013

A critical step within multidisciplinary collaborative work involves establishing a consistent group vocabulary or taxonomy to name what we are talking about.  This must be followed by establishing agreement within the group about the term and its meaning.


We are currently working on framing the “story” we want to tell about the choreographers’ work.  As part of that process we have begun to identify elements that we observe in the choreographic ideas of Bebe and Thomas’s work.  This naming of elements is to help us to communicate within our working group.  Some of the terms come through in discussion and some were pointed out and given terms by the choreographers and/or their dancers.  In the initial stage we are concerned with the naming and visual identification. We will eventually consider how the vocabulary will communicate with the intended audience for the project.


In the Synchronous Objects project a long history of a vocabulary specific to One Flat Thing, reproduced existed within the company.  This is typical for the way dances come to be.  Elements get named with an insider vocabulary that helps the dancer and choreographer to be able to recall it and reference it.  This vocabulary can be expanded as the choreography lives on and the dance is passed through different dancers who learn the piece at different times.


Differing vocabularies have long been a stumbling block in collaborative, multidisciplinary teams where people tend to use different terms to describe a similar concept, depending on their backgrounds, training and experiences.  And because we need to be able to discuss the work at the originators level, at the team project level and at the intended audience level we are now in the process of creating a vocabulary and its definitions that will enhance our communication about these choreographic methods that we are focused on illuminating.


The process looks like this:

Identify “things” through discussion and observation Get vocabulary from the originators and work with movement analysis experts on naming and term conventions compile the vocabulary and identify and consolidate differing terms for like meanings*  Return to originators to confirm the “things” and their meanings  find visual examples of “things” to confirm that we share the same meanings  formulate ideas for new names and attach text definitions and visual examples to each  share new names, definitions and visual examples with originators and project team  refine naming and definitions based on discussions with originators and project team  finalize vocabulary to share with intended audience demographic for user testing  refine naming and definitions based on discussions with intended audience demographic  share changes with originators and project team  accept final vocabulary and definitions.

* indicates our current stage


Having used this approach in prior multidisciplinary projects we have come to recognize it as an essential part of multidisciplinary collaboration.  I am currently working on another project with colleagues from Nursing and Engineering and the same need arises during different stages of the process.

– Maria

Animation and Choreography in the classroom

Posted on June 18th, 2013

The relationship between choreography and animation is symbiotic. Each time Norah and I collaborate across our disciplines, we reaffirm the desire to construct a course that explores both choreographic and animation approaches to movement, by integrating both dancers and animators. (I don’t know of any example of such a course existing in college curricula, but if you have knowledge about it – please send us a note.)


This Motion Bank project has inspired us to try an exercise in getting animators and dancers working together. Using our two co-scheduled Spring 2013 courses (Expressive Animation and Interdisciplinary Creative Research Seminar), we designed an experience for our animation and dance students that engaged them with Thomas Hauert’s Careful Scientist exercise.


We began by introducing the students to Thomas’s exercise by looking at a video of his performance.  In the Careful Scientist exercise, Hauert defines a limited set of instructions or rules for moving the body. He defines the rules of the system in such a way that they challenge mind/body habits.  In order for the students to better understand the work, concentration and coordination necessary for this exercise we had them practice some of the basic concepts by trying these movements in groups with others assisting.  Next we presented the students with some of the data of the Careful Scientist that we had motion captured during one of Thomas’s residencies at ACCAD. This provided a rich data resource of 3D positions and timings that describe the movement.  We then challenged the students to reuse existing motion captured movement data to re-imagine and design a new system of relationships, driven by existing movement data to become a new movement phenomenon. Three outcomes from that assignment follow,  accompanied by the students’ explanations.


Flowers by Maddy Varner and Daniel Diller

Concept: “We are using the data of an unnatural movement process and placing it on a natural environment. We are creating movement in the animation that is driven by external forces (like wind) while the careful scientist is propelled by internal choices. “

Tech Notes: “Upper Arm Pointing data occurs on the X and Z axes of the shoulder joint and is assigned to the direction of the flower stems movement (making them look like they’re being blown by the wind). Shoulder Rotation data occurs on the Y-axis of the shoulder joint and is assigned to the wilting and blossoming of the flower heads. Elbow Flexion data occurs on the Z and X axes of the elbow joint and is assigned to the growth of the flowers. Forearm Rotation data occurs on the Y-axis of the elbow joint and controls the color of the flowers. “


Spork! by Jonathan Welch

Concept: “I used a motion capture of the Careful Scientist by Thomas Hauert to create a visual parody on the hybrid theme prevalent in art and technology work. All the motion is a combination of the Careful Scientist, at different speeds and directions, with altered key frames to influence the gestures and correct for problems translating a human’s motion to the altered anatomy of the characters. The combined motion is intended to be dis-articulated, awkward, and hyper extended; alluding to a dysfunctional parental couple whose behavior is compounded through their child. “
Tech notes: “To start, I sped up the motion capture data, duplicated it, translated it forward, and attached both to the spoon. Then I duplicated the motion again, and broke it down into key frames in places where the hand went through the head and moments that I decided to make the existing gestures more like an argument. I wanted the fork to have a slightly different personality, so I took the key frame motion path, duplicated it, reversed it and attached the motion to the fork. The spork’s motion is the same as the spoon, without the argument key frames, and not quite as fast. “


Ontology of Thomas Hauert’s ‘Careful Scientist‘: Judith Butler’s reading of Derridean “Citationality” by Carolin Scheler and Kaustavi Sarkar

Concept:  “Dance scholar Susan Leigh Foster describes improvisation in dance as ‘bodyful’ dwelling in the perceptual gap of the conscious and the unconscious. Continuing the quest of the unknown through digital inscription of dance, we argue that digital ontology exists in corporeal materiality of the performing body as performatively, affectively and digitally mediated in the animation of Hauert’s Careful Scientist. Through Butler’s reading of ‘citationality’, in which ‘matter is constructed not as a site but as a process of materialization that stabilizes over time to produce the effect of boundary, fixity, and surface we call matter’, the digital is ontologized in Hauert’s performativity.”

Tech Notes:

Diagram to explain motion capture data mapping to 3D geometry

Diagram of motion capture data mapping to 3D geometry

ACCAD animation specialist and motion capture expert Vita Berezina-Blackburn has continuously pushed our Center in thinking about unconventional ways to use captured motion. As an animation artist and collaborator on Bebe Miller’s Landing Place her work demonstrated the aesthetic beauty that can result from expertly combining captured motion and form. Take a look at some of Vita’s work from Landing Place.


– Maria


Posted on June 17th, 2013

The Careful Scientist exercise requires following a small set of joint movement rules which dictate the ways a performer’s arms may move.  As part of the rule set no movement that is happening on the one side of the body can happen at the same time on the other side of the body.  Because the options are so discrete and constrained, the number of possible action sequences can be calculated.   As we have discussed here previously, the movements are limited to:

  1. Upper Arm Pointing
  2. Shoulder Rotation
  3. Elbow Flexion
  4. Forearm Rotation

With a new action being chosen roughly every 1-3 seconds, there are millions of possible sequences for the first half minute of the exercise. Calculating the number of permutations was useful for giving a sense of the number of possibilities for this seemingly very constrained exercise. It also furthered discussion of choice-centric vs. action-centric perspectives of the exercise.  Diagramming the options for the beginning sequence, and the possible choices that emerge quickly revealed the simple pattern that follows.


From an internal, experiential choice-centric perspective, I note that I begin by choosing a starting action to perform in one of my arms (“action 1″) followed by a starting action to perform with my other arm (“action 2″). I then begin moving by starting both action 1 and action 2 at the same time.  From this perspective, I had 8 possibilities for my first choice.  My selection then led to only 3 options for my second choice.  There are 8*3=24 possibilities for this initial sequence of choices.


However, from an action-centric perspective, when I simultaneously begin performing my initial choices, “left Elbow Flex & right Forearm Rotation” is exactly the same activity as “right Forearm Rotation & left Elbow Flex”, regardless of the order in which I selected these two actions. From this perspective, there are (4*4)-4=12 initial combinations for how the arms can begin moving. (Four possible actions in each arm yields 16 possible combinations, minus the four “illegal” combinations where both arms are doing the same action.)


Left   Right

1      2,3,4

2      1,3,4

3      1,2,4

4      1,2,3


After both arms begin moving, as long as you “change the actions in both arms at different times, so that the action in one arm continues while you change action in the other” (Thomas) then each additional change multiplies the permutation count by 4. This results from having four options at any time for the next action: the left arm could change to one of two other actions, or the right arm could change to one of two other actions. Neither arm is allowed to repeat their current action, and neither change to the action the other arm is currently doing.


action choice permutations

1      12     12

2      4      48

3      4      192

4      4      768

5      4      3072

6      4      12288

7      4      49152

8      4      196608

9      4      786432

10     4      3145728

For ‘n’ actions: P(n) = 3 * 4^n


While the above shows the action-centric perspective, the only change necessary to switch to the choice-centric perspective, is that actions 1 and 2 have 8 and 3 choices respectively. The equation changes to P(n) = 6 * 4^(n-1), for n>=2.


Next steps could involve determining how variants of the rules effect the number of options. For example, if you allow both arms to change their actions simultaneously, how does this change the number of permutations? I suspect this allows the arms to swap actions, potentially increasing ‘4’ options to ‘6’, but there are dependencies that require further consideration.


I wrote a script that counts actions, pairs of actions, and trios of actions for each (and both) arms.  One of the next steps may be to generate some graphics to reveal patterns or lack of patterns in their action choices in pairs and trios.


And graduate research assistant J Eisenmann pointed out that there are some other ways to show the action ordering frequencies.



-Matthew Lewis, ACCAD Graphics Researcher