I'm looking forward to seeing new music gear at the Audio Engineering Society (AES) convention in San Francisco this afternoon. However, I doubt I will see anything quite as innovative as this musical robot:
ATLANTA, Georgia (CNN) — A professor of musical technology at Georgia Tech, Gil Weinberg, enlisted the support of graduate student Scott Driscoll to create Haile — the first truly robotic musician…
…Think of Haile (pronounced Hi-lee) as a robotic partner in the percussion form of dueling banjos. Although it has numerous musical algorithms programmed into it, Haile's basic function is to “listen” to what musicians are playing and play along with them…
…The robotic drummer is not only programmed with specific pieces but also with an understanding of countless pitches, rhythms and patterns, which are used during performances. Like a concert drum solo, Haile never quite plays the same thing twice, but plays off the creations of those performing around it.
We'll know we have truly created robotic musicians when they show up late to gigs and step into the alley between sets for a smoking break and other recreational intakes.
Actually, I did some research into computer programs that can “listen” to music and generate new material that is similar but not identical, way back in high school in 1990. Basically, it recorded MIDI input, created a decision tree of sorts, and generated new music from it. I had good time working on that project, and while my teachers and a few computer-music faculty I talked to at MIT and elsewhere seemed impressed, other “distinguished scientists” brought in to review our research projects largely pooh-pooh'ed this work as “not being science.” True, it was not a science experimental per se, but I've always been bothered by those in academic science that don't value innovative engineering. This happened in graduate school as well, where the computer science faculty wanted “science.” I wonder how they would have reacted to Weinberg's musical-robot project? I have always been more interested in making things than taking measurents and analyzing data, and thus have been mostly turned off from pursuing a career in scientific research.