Introduction | Background Work | Expressive Processing | Conclusions and Future Work | Bibliography

Conclusions and Future Work

With the advent of high performance computer systems it became possible to use real-time processing technology as common practice in Performative Arts.
Even though digital technology allows very high data resolution and detailed analysis, in many cases superior to the human perception, it is also truth that algorithmic cognition and interpretation processes are far from reaching the human standards.
Specifically in the case of gesture capture and analysis, even though existing research has provided very powerful and open tools, such as the Eyesweb Software, existing systems are mostly effective in the specific aspect that drove its development.
This is the reason why in Art projects, development should depart from ideas and human concepts and implementation should be at the lowest technological level possible. This is the only way towards overcoming the constraints imposed by existing tools.
In fact, this is one of the main directions in future work for “Expressive Processing”. In future developments of this project, the basic system based on Eyesweb and Virtools applications communicating trough MIDI should be ported to custom developed software for computer vision and image synthesis in order to achieve a higher detail in the articulation, smoothness and responsiveness of the system.
This can take the dancer to the next level improving interaction facets and defining the constraints imposed to her performance with more detail and coherence with the author’s ideas.
At the post production level, for this specific piece, the use of Max/Jitter paradigm in algorithmic editing based on MIDI data representation of the dancer’s movement (12 channels corresponding to the relevant body joints captured in EyesWeb) has reveled sufficient and effective.
Further future work in this piece will include more interaction with the music composer at the post production level and the use of better sound generation modules (the piece will be recorded in a MIDI controlled acoustic Yamaha Disklavier Piano).
At the musical level it is also possible to use MIDI data from the dancer’s movement in the horizontal axis to generate a sound spatilization matrix for reproduction in DVD format (surround 5.1).
Overall this work intends to demonstrate that following the transversal concept of “constrained expressiveness” over different setups, diverse results can be achieved, but the final outcome is extremely fruitful and coherent.
In all the pieces it is clear that dancers can adapt to the conditions imposed by the systems and produce an expressive and rich performance.
Specifically, in “Expressive Processing”, the use of technology to register a performative dance event at the visual level in synchronization with gesture and motion capture logic data, has reveled to be a powerful paradigm to generate and process a post-produced video based in the interdependency of these two data sources.
In the final video the music is made from the sonification of video data and video editing is generated from the dancer’s movement and gesture transcription to MIDI.
Therefore, it resulted in a rich and expressive audiovisual piece that unveils the expressiveness behind the synchronous bounds between the video, music and image composition.

| A Project by Helena Figueiredo: hfortuna[at]final.upf.es; hfigueiredo[at]porto.ucp.pt |