It is possible to sample pixel data and get the color values with the pix_data object. It accepts a normalized x, y position (ie, between 0, 1) and outputs the color and grayscale value of a pixel. You can use this to drive audio if you want.

Later, Mark

============================ = mdanks@stormfront.com = Lead Programmer PS2 = http://www.danks.org/mark ============================

-----Original Message----- From: Michael Lechasseur [mailto:mlecha@music.mcgill.ca] Sent: Tuesday, August 01, 2000 8:10 AM To: Pd-List Cc: Michael Lechasseur Subject: images -> audio

A visual artist would like me to create some audio from digitized brain scan images. I peeked at the objects in the GEM library, and they seem to deal with creating and displaying graphics, rather than using graphics as a source for output to "pure data".

I seem to remember a friend telling me that pd and GEM were used at a demonstration at McGill university, and the patch was programmed to created audio from a graphic file.

The processing of the data used to create the images would be fairly minimal.

Any suggestions?

Michael Lechasseur