patch, software and hardware tests with glued videos distributed in matrix. anticipated with various different videos distributed in matrix and same reaction clip for all, however had to be reduced to only one video multiplied in matrix with same reaction clip. different video inputs and face recognition trigger to different/reaction video input to some parts of the video output.
tests made with different matrix settings, various video input and reaction clips.
specifications: Max/MSP 7, FaceOSC, Camera, 3 Screen Set up with Matrox Triple head. (for projection need to use video mapping extensions, f.i. MadMapper)
several issues: processor capacity to play multiple different video files at a time and in a glued matrix whilst running face recognition triggerresolution; low resolution video only; differences of resolutions and video codecs for glued matrix.
lessons and steps to look into: min. 2 or 3 computers needed, with better processors and video capacity. how to play high res video input at the same time and in matrix, potentially feed via VLC/Quicktime and Madmapper instead of gluedmatrix. use maxhole to send input from diff computers on same network. need for multiple cameras or wide angle cam, or kinect with synapse or kinect eye to recognize the exact position of ppl instead of eye/faces. also, design/ video clip wise, and reaction clip, some further variations need to be created as it is still suboptimal, and does not create the desired vis effect.
