|
nime: new interfaces for musical expression aya karpinska other journals: one || two || three|| four || five || six || seven || eight || nine || ten (you're here) |
|
|| journal ten|| 29. IV. 03 || a spasm, a lull, a time for performance. i won't jinx myself and claim that i have the final version of the BX-24 code, but this seems to be working well: MIDIMax2.txt. i was sending five different control messages, with a delay of .2 seconds between each. that doesn't seem like much, but by the time i get around to sampling my first value again, a whole second has passed by - the latency was definitely noticeable. after i took out the delay, the project simply didn't work. terrible. after stressing all afternoon, i found out that i had to increase the size of my output buffer so that the MIDI receiving module wouldn't get overloaded. i added a short delay to the end of my loop (0.05 seconds) and that was it. things are working well. my Max patch has improved as well. i now use the switches to change patches, and figured out an algorithm for what i call "hold until change". there should be a Max object for this, but i couldn't find it. basically, i want to be able to only pass through a value if it changed. normally, you can use the "change" object for this, but i wanted the value passed through once and only once. so the pseudocode is as follows: "if integer 1 is not equal to integer 2, then pass integer 1" which in Max is written: if $i1 != $i2 then $i1 thanks to aaron miller of buffalo, ny for his help on that one. we met at a festival of digital poetry this weekend, and after his Max/Jitter-based real-time performance i seized his arm and convinced him to help me think through it. spread the knowledge. i'm very satisfied with my manipultation of samples. i have two basic patches for now, and am considering adding a third. one is very electronics-sounding, kind of subtle and soft. the second is full of squeaky sinusoid madness. (i swear, these description make sense to me...) i have to think more about how i will play the two during performance. right now i plan to start and end with the more minimal one, with distinct pauses between patch changes. that way i don't overload the system - crash during live performance = bad. i don't often exceed 40% CPU usage according to MAx's DSP status window, but i'd like to keep things safe and not too processor-heavy. i loved messing around with the box today, and look forward to more improv later in the week. time for sleep now, i've had too little sleep for too long.
|