awesome combination of improvisation performance, rhythm, percussion, and robotics! "One human, three machines, rhythm,"is what is written on the video description, as well as what they are coining the "steamfunk" movement.
quoted from Patrick Flanagan:
There are two modes of performance with Jazari; I call them "performer" and "conductor." In conductor mode, I control high level parameters, such as loudness, phase-shifting, degree of syncopation, and M-Operation. (Doc 02, I'd say that live control over the latter parameters *is* new). In the video, I don't use this mode at all. I do trigger the loops in the cabasa, cowbell, and clave, but that is the only pre-sequenced material.
The rest of the notes-99% of everything played by the Djembe and bongos-are produced in performer mode. In this mode, I control the choice of individual notes-just like a traditional musician. To Feral, I agree with you about the drawbacks of track-based musicians; that's why I improvised almost all of what you heard on a note-to-note basis (I also agree about temporal quantization; a model of expressive timing is on the to-do-list; velocity is already variable).
I'm able to improvise these fast, complex passages because my software interprets holding down a button as a decision to repeat the previous note. That doesn't make playing these things easy exactly-they present their own challenges-but I was able to get to this level with no percussion training and six months of regular practice.
In this project, I trade-off a degree of expressive subtlety for improvisational power (I'm playing two instruments at once throughout). Beyond juggling multiple instruments, the new kinds of idiomatic playing opened up by the controller interface and the possibilities of self-refracting man-machine interaction make the trade-off more than worthwhile.
