Sunday, October 7, 2007

POLYTOPIA questions

This is the cover art of my new CD, POLYTOPIA which came out from Bridge Records in September. I love it and so grateful my mother in Japan didn't hate it! (in fact she LOVED it, too)

Anyway, many have asked about my title piece of this CD, POLYTOPIA for violin and signal processing and how it was done, so here are some pointers. I am using an interactive computer music program MaxMSP ( which has been the main tool I have been using for more than a decade. I basically only use two techniques in this piece; pitch shifting and delay. I am using SIX independently controlled pitch-shifters in MaxMSP (harmv2~) and delays, and also they are independently panned, as it was originally (still is) conceived as a Surround 5.1 piece. My visual image was that if it were a real-life situation, 6 vioinists will have to be running across the hall really fast without tripping over, while playing like mad: it's kind of a funny picture!

If you would like to know more about nuts and bolts, here are some nitty-gritty details:
When there are chords in unison, I make sure that the 6 voices are not exactly aligned--I make sure that they are about 5-50 milliseconds apart, as it would add a 'fuller' or 'real' feel as if those chords are played by humans. No REAL six human players will ever exactly come in absolutely the same time, zero milliseconds apart! I also make sure that these 'artificial delays' in unison is not always the same--i.e., if delays on each voices are: voice#1 = 0ms, #2 = 10ms, #3 = 50ms, #4 = 30ms so on, then the next chords I switch the delay times to 20, 36, 5, 54ms so on, so the chords don't sound uniform.

I had also made my own tuning table for my pitch-shifting, thinking that the string tuning is very different from equal temperatment--so I built my own pitch-shifting intervals by ear. Then, I listened to the result and it wasn't as 'beautiful' as I thought it was going to be, aside from the fact that I am not staying in one key (which would have made sense). So, I bravely discarded my 'beautiful' tuning table which took HOURS to build, and went with equal temperament tuning (1200 cents devided by 12). I often find that even in acoustic composition, just because you spend many hours on something, doesn't mean the result is good, and I need my courage to say NO :)

In my opinion, the 'death' of computer music can be easily achieved by your LAZINESS :) I try to vary as much details as I possibly can, since I think audience is acutely aware and sensitive to monotony---if they catch on something that is exactly the same, you would be surprised how fast you can lose their interest. You have to keep them wondering---besides if the details vary you can also react more musically, as well as amuse yourself and keep yourself fresh :)

Time to read some Japanese books for my son since he brought me one... a GOLDEN opportunity for a poor boy whose parents are trying to raise him trilingual... more later...


Wendi said...

Why do you work with MaxMSP?

Mari Kimura said...

Hi Wendi,

Thanks for visiting!
I use MaxMSP because it is the only program that allows me to truly interact live with a computer, that is commercially available and relatively easy to use; the initial learning curve maybe steep, but even non-technical person such as myself can teach oneself to use it. MaxMSP is the most versatile program that exists today for the public--there are many people in academic institutions using their own systems probably better than MaxMSP, but they aren't commercially available. Other commercial programs such as LIVE or Logic could do certain things live, if you pre-program it, but you can't interact such as using pitch tracking, loudness tracking and much more. It all depends on what you want to do musically, and it might turn out that you don't need MaxMSP. The next best option, if you are processing, I would choose LIVE, or even Logic, and automate effect change, to which you can 'play along'. It is akin to Karaoke, except the karaoke part is the processing :)