Computer Choreography / Choreology Project



The programs and projects described here are no projects of Universität Frankfurt. I could not foresee if and when the described programs will be finished. I have not decided jet if the described programs will be publicly accessible in any form.


This hypertext document contains a short description of my current project for Labanotation and the simulation of dancers. Labanotation is a standardised system to analyse and record human motion. Mainly it is used (similar to music notation) at theatres to archive ballets. The input of this notation into the computer and the decoding of it into a 3-D computer simulation of human motion is the goal of my project. Based on this project the implementation of a expert system for choreography is possible. A system who enables everyone to choreograph.


Choreography, Human-Simulation, Labanotation, Dance, Handwriting-Recognition.


1. Preface
2. Introduction
3. Why using Labanotation?
4. LabanPad shortcuts
5. Limelight shortcuts
6. NKvsMK shortcuts
7. Epilogue
8. Links to additional Pages concerning this Projects

1. Preface

My primary goal is currently to realize a Choreography of my own (11 Min. - classical ballet for 22 dancers). I don't have the necessary dancers therefore i have to do a simulation of the missing dancers. The programs described here are only tools to do this. The essential of dance is that dancing is done by humans, in simulation dance losses that what is it's essential. If my project is finished the state of choreography could never be the same. The access to choreography and the necessary dancers has changed from a rare good to a free accessible good. What dance wins and losses through this change is obviously. Dance will be more democratic but the essence of dance is missing. Maybe more quantity produces particularly more quality. If it's possible to write a expert system for choreography then the question: "is choreography teachable?" is solved. The quantity and the way how computer choreographies are build (guided by a expert system) make it impossible to judge the resulting works.

2. Introduction

Human simulation is currently one of the main research fields in computer science. Future applications of the results of this research affect medicine, engineering and the performing arts. My project is a concrete application of this newly developed technics in the field of synthetic actors.
Most the time choreographers don't have access to the necessary dancers who they need to realize their ideas (especially in the education of choreographers). Simulated dancers are available in a unlimited number, they have unlimited patience, they are good in memory, they don't need repetition to remember something and it's easy for them to forget.

3. Why using Labanotation?

Computer scientists tent to reinvent the wheel. In the current research in human simulation they are to much struggling with basic technical problems, instead of improving the usability of their systems. The anatomic models are most the time to simple - and almost nobody thinks about how to input the movement sequences.
The three current approaches to this problem are so far:
1. Form the figures to the end position of the movement (keyframe animation).
2. Input of the movements in natural language.
3. Sampling of the movements (motion tracking is not simulation!)
Keyframe animation is useless for large scale animation. Natural language is interesting for some simple situations. For choreography natural language input is useless by its leak of precision and its difficulty to survey. Using sampling as input for a choreography system is like if you try to write a program for composing, and let the composer input his symphonies by whistling.

The research team of Norman Badler investigated in the early days of their system Jack in the use of Labanotation for motion input. Unfortunately they don't continued on this path - the effort of learning Labanotation seemed to much for the scientists. If they had used Labanotation in this days, we probably had today a working system for choreography.

Labanotation has the advantage of existing now for 70 years, and being further developed in this years. Until now thousands of pieces have bee recorded in this language. Labanotation is through its column structure very computer like, through its visuality very easy to learn, through its parallelism very survey. Therefore is Labanotation contend of LabanPad, main medium in Limelight and in NKvsMK the internal substructure.

4. LabanPad shortcuts

LabanPad is a program for interactive Labanotation input. It will be implemented on a pen-based palmtop-computer (Currently: Apple Newton), which is usable "on location". It is based on a quite natural user interface, the input of Labanotation is done by writing the symbols onto the display, a specialised handwriting recognition algorithm is used to convert the strokes to a tokenised code. The editing of the notated material is done by gestures like mark and cross out. The goal is to substitute pen and paper by a almost as easy to use but a quite more powerful tool, which is usable in the ballet-room in the hand of the choreolgist.
-> More information about LabanPad

5. Limelight shortcuts

Limelight is a modular program which integrates all components, who are necessary for a professional computer aided choreographing. It's the host program of LabanPad and it should enable the user to work in a project oriented way. All media are collected in a project and can be synchronised. The the media ranges from a storyboard, music (as sample and as notation), video, pictures to animation control data. A major instrument is Labanotation, out of the notation a 3-D computer animation will be generated. Labanotation will build also the basis for a expert system for choreography.
-> More information about Limelight

6. NKvsMK shortcuts

While LabanPad and Limelight are tools for the professional user - the choreologist and the choreographer, NKvsMK is just a toy. It should produce the illusion the individual user could choreograph. The program will guide the user through a always limited number of selections inside a path calculated by a expert system for choreography. This program is in some way a Trojan horse inside a violent computer game scene, which should replace destruction with aesthetics and creative work. The new 3-D game sets like Sony-Playstation, Sega-Saturn and Nintendo-64 will make such a program possible.

7. Epilogue

The preparation of my projects are now finished. I'm implementing Limelight. After all you should not count on a fast succeed of my projects because it's a "High Tech, High End, No Budget - Project" (If you know any friendly "Founder" who likes to found the work of a dozen programmers in India ...). Well, i need this program (And i don't have to pay my own work). The time is right, the (computer-) technics have reached the necessary level and are soon affordable, lets try it again in a new way!
I'm always grateful for hints or suggestions, especially hints to similar projects - maybe i could save a lot of time and work using already existing programs to realise my choreography.

8. Additional pages concerning my projects

Information about LabanPad
Information about Limelight
Information about my handwriting recognition
Introduction to Labanotation
Links and books about the topic
Information about me
My Homepage

Uni - Frankfurt Homepage

(C) 1996 Christian Griesbeck, Frankfurt/M
mail to: