testing
I’ve pretty much finished the final 4D project in VVVV now. Reacting to motion the portrait shaped screen render enlarges a grid of blocks and, depending on how erratic the movement is, shades them red the larger the frame count goes.
Some tests from various places and distances have proved very good. The fourth floor of Babbage during the day proved a little too tricky for the web-cam I’m using but my bedroom window at night (sounds creepy) and across a 20 metre room proved to work really well especially to vehicles at night. I’m hoping that the distance of third floor Portland square wont be too far for the cam and that the human and traffic movement will be frequent enough to create a lot of movement from the screen.
The VVVV patch I’ve programmed for this uses a motion capture node which measures when a new frame is created and the old frame is deleted virtually every second, or at least every time it picks up movement.
This motion capture node is connected half way between the web-cam detection node and the web-cam render window and can be enabled or disabled and then this feeds into a series of nodes that render the block grid and alter its size and colour finally feeding into the block render window.
This allows for a completely real time reaction to live movement as and when it happens. As far as I know there hasn’t been a node developed or VVVV yet that maps the screen area in which the movement occurred. This would have added a gradient effect to the block render however because of the low resolution nature of the green screen the colour block grid is very effective.
