Final Project :: openCV + OSC to MaxMSP

for my final project, i used the OSC library to send messages to MaxMSP to control sound synthesis and sampling. My concept was to create a physical environment for someone who has a lot of control over their body to be able to move through, and have legitimate control of sound. Someone like a dancer could potentially see this as an instrument. In Max, I parsed the OSC data to read the newest cell with an "energy" (CV detection) greater than 0.5 and triggered one of 3 sound patches. I'm using FM synthesis, and two types of granular synthesis to create my sounds. If ever implemented in a big scale, I would have two cameras, perpendicular to each other, in order to create a full 3-D cube in which the dancer would perform inside. This would give all 3 axis' different control of various parameters of the sound. I thought the idea of breaking the CV image into a matrix of cells was effective because without these "trigger" areas, the user would constantly hear change, even as they try to stand still. Here, one can actually aim for different notes or areas, and stand there without triggering extraneous sounds. In a performance setting, there would have to be someone "mixing" the sounds as you can see me doing in this vid - switching between what patches are getting triggered. The results are mixed in terms of control, but I was successful in laying down the foundation of what could be what i think is a really cool performance.




Here is the source and maxMSP patch. SOURCE.




homework 9 :: graphics + synthesis

part a • simple polygon + sinusoid FM synthesis




part b • granular matrix







homework 8 :: white squares

part a • FFT peak analyzer




part b • aubio pitch and amplitude analyzer








homework 7 :: sound from image and image from sound

to begin i composed this, and passed it on...




then i received this, from zachary gage...




and made this visual transcription - to be read left to right, top to bottom




i also received this graphical notation from zachary:




and went ahead and composed this track based on that






homework 5 :: • GRANULAR SYNTHESIS •
recreate sounds, granular image, gestural grains



GranniStation • gestural control of granular synthesis from j saavedra on Vimeo.

Part 1:: granularize an image

••• all images have been generated using the same source image.
click assignment 5 to see source •••

grainularized1::


grainularized2::


grainularized3::


grainularized4::


grainularized5::



Part 2:: mimic sounds w/granular synthesis


"wind blowing past your window"



"tea pot, ready to explode"



"bubbling brooke, underwater"






homework 4 :: non-linear sequencer, waveshaper

Synthicles v0.1 - me, bruce, nick, jen


in it's current state, there is very little control over the balls. hopefully we will return to this, and make axis restrictions possible for each ball (instrument). Visually i am very happy, but as a musician primarily, i want more control. I would never use this to perform in it's current state. still, a great experiment



part b: waveshaper


512 random values are generated, and stored in an array. the values are then mapped directly to affect the carrier frequency. both carrier and modified waveform are shown in this video.






homework 3 :: create a sequencer


part a • nick hardeman, jen cotton, me - sinusoid step sequencer v04







homework 2 :: story told through sound

"chicken and lamb on rice"






homework 1 :: from image to sound